Bad Line-up?
  News    Contact Us    Links    295 Search    RailCity    VRU    
  News & Updates  
  News Archives  
  Agreements & Benefits  
  About the TCRC & 295  
  Resource Library  
  Rights, Safety & Health  

Have you been forced to be on duty over 10, 11 or 12? If so we need to know why.
Click here...

Railroad workers have been fighting fatigue in the rail industry for decades but the problem persists.
Click here...

295 Meetings
The next meeting of Division 295 will be held on March 14th at 19:00.

Map... | Meeting Schedule... 


Current status of fatigue countermeasures in the rail industry

bullet Introduction
bullet The Science of Sleep and Fatigue
bullet Effects of Fatigue on Performance
bullet Technological Countermeasures
bullet Past and Current Developments
bullet Conclusions

Source: School of Education - University of Denver

The study of the role of fatigue and transportation has a long history. As early as 1917 a task force of scientists was gathered to study the effects of fatigue on drivers accidents. Due to the ever changing complexity of the demands faced by drivers and operators in all modes of the transportation industry this topic continues to be the focus of intense study (Sherry, Bart, & Atwater, 1997). Over the past few years there have been increased efforts to address the problems of fatigue in the railroad industry. A USDOT/FRA report in 1991 (Pollard, 1991) identified causes of fatigue. These are: uncertainty about the time of one’s next assignment, excessive working hours, long commutes and waiting times before beginning work, unsatisfactory conditions for sleeping at some terminals, and the decision not to rest during the day even when subject to call the next night. Suggestions for remedying the situation included: a minimum of eight hours notice before being called to work, greater predictability in scheduling trains, division of assignments according to blocks of time.

A later report by the GAO (GAO, 1992) found that railroads were essentially complying with the Hours of Service Act. In fact, it was found that 99.4 % of the time engineers were given at least 10 hours off duty following a work period of 12 or more hours. Further, the investigators found no instances in which an engineer received less than 8 hours off duty in any 24-hour period. It was also found that engineers rarely worked more than two consecutive shifts with fewer than 9 hours off duty between shifts. The report indicated that reducing the maximum number of hours allowed per shift from 12 to 10 would have little effect on the number of accidents that occur. It was found that only 4.5% of all human factors caused accidents occurred after 10 hours in an engineer’s shift. The report further cautioned that reducing the "maximum allowable work/off-duty periods from the current 12 hours on, 10 hours off cycle to a 10-on, 10-off cycle could increase the variability – the change in work period start times from day to day – of engineers work cycles." The report cited research that suggested that variability in work cycle start times disrupt natural human sleep-wake cycles, which in turn can lead to fatigue.

The GAO study found that more human factor caused rail accidents occurred from 2 a.m. to 6 a.m. than in any other 4-hour segment. Incidentally, the accident rate between 2 a.m. and 6 a.m. was higher than at other times. "The start time variability of engineers work cycles was quite pronounced during the 2 a.m. to 6 a.m. time period." The report authors speculated that, " Higher levels of start time variability increase the likelihood that engineers will experience fatigue." (GAO, 1992, p. 3)

A study (Kuehn, 1992) of four engineers observed under a normal and a fatigue work schedule, concluded that a deterioration in engineer performance, regardless of schedule, coupled with the irregular sleep/work patterns of the subjects suggest the need for continued research which focuses on sleep work patterns and performance. While as a group the study participants did not differ in overall performance they were observed to incur speed limit infractions, failures to blow the horn for crossings, rapid throttle changes, and application of excessive train forces. Thus, specific instances of performance decrements were observed.

This study was followed up by a more recent study of 55 engineers monitored while operating on two different railroad work schedules (Thomas, Raslear, and Kuehn, 1997). The first schedule was designed to run "faster" than another group running "slower" in terms of frequency of train operations. Engineers operated trains in a simulator for a ten-hour shift. Participants had at least an average of 9.3 hours off duty for the "fast" group and 12 hours off duty for the "slow" group. Results showed that the "slow" group got about 6.1 hours of sleep per night compared to 4.6 hours for the "fast". Performance measures included number of missed horns sounded at crossings and cumulative pounds of fuel used. Results showed that the "fast" group missed about one third more horns at crossings than did the "slow" group. Furthermore, the "fast" group used about 200 pounds more fuel per trip segment than did the "slow" group.

In 1995 the NTSB and NASA in cooperation with the Department of Transportation sponsored a symposium on fatigue. The conference was thought to be a first step in educating diverse groups in the transportation industry to address the fatigue issue. Dr. Mark Rosekind, a presenter at the conference summed up the intent of the proceedings in his introduction: "An important theme expressed throughout the entire symposium was that there is no magic bullet to eliminate human fatigue in transportation operations…. Every participant is encouraged to take some action to educate, address a scheduling issue, use a countermeasure or apply some piece of knowledge.... to improve transportation safety." (NTSB, 1995).

Given the previous work in this area and the current legislative climate it was decided that a detailed study of the current status of fatigue countermeasures was needed. The purpose of this monograph is to document the nature of these interventions and to comment on their credibility from a scientific point of view. In order to make these comments it will be necessary to first review general principles of fatigue countermeasures. This will be followed by a discussion of individual railroad programs.

The Science of Sleep and Fatigue
Source: School of Education - University of Denver

In depth study of human fatigue has only occurred during the last twenty-five years. The scientific community has discovered the structure of sleep, the role of circadian rhythms, and the effects of sleep deprivation. Researchers have also investigated the benefits of sleep management including a napping study by Dinges (1995).

Before entering into a discussion of what is known about the science of fatigue it is important to realize that at this point while much is known, much is not known. A recent book summarizing the effects of sleep and performance cited over 400 scientific studies investigating the topic. Many of these studies examined the effects of sleep, sleep deprivation, napping, and alertness on performance. However, the results of these studies do not provide practitioners and employers with a definite "answer" or "specific treatment" for ways to minimize or eliminate the effects of fatigue on performance. Science progresses in an incremental fashion where established ideas, or "common sense", are subjected to rigorous scrutiny and testing. This rigorous process allows others, anywhere in the world, to replicate the procedures, methods, and analyses to verify for themselves the findings that have been reported. Not surprisingly, however, different experiments in different settings under different conditions studying the same phenomena can sometimes produce apparently contradictory results. This is often the case in new areas of inquiry. The study of sleep is no exception. As will be seen there are some findings that have raised additional questions.

While the incremental increase in our knowledge progresses steadily, many questions remain. This is due in part to the fact that scientific studies are conducted under specific circumstances and conditions. As studies accumulate there are findings that appear to be consistent with each other and some findings that are anomalous. The scientific community attempts to conduct studies that replicate some portion of a previous study under new and different circumstance. While some results may be similar to previous studies, other results are not. Thus, in science it is normal to expect that there are inconsistencies in the findings of various studies as scientists attempt to discover the underlying laws and properties of particular phenomena they are investigating. Anomalies are the driving force of science, things that we can not understand or explain drive the field forward. However, for the lay person these inconsistent results may be disconcerting and pose challenges for practical application.

Thus, while inconsistencies in results may be useful to scientists, the extent to which they are generalizable to everyday life may vary depending upon the degree of similarity to actual real world conditions. Moreover, while several studies have produced clear results in the laboratory, other findings may be less readily transferred to an actual work place environment.

One area of confusion is the way to measure fatigue in the workplace. Many studies rely on several different approaches to determine the effects of fatigue countermeasures on performance and alertness. It is common to use both subjective and objective measures of fatigue and alertness to evaluate the impact of a countermeasure. Multiple measures allow the investigator to triangulate the truth and produce a more convincing conclusion, as more information becomes available. For example, a study by Johnson et al. (1988) used both subjective and objective measures to assess sleepiness. A correlation between the two subjective measures was significant at .52. However, the correlation between the physiological and the subjective measures was only .18, a non-significant finding. A later study by Kecklund and Akerstedt (1991) found a significant correlation between subjective measures of fatigue and EEG alpha burst activity at .38. The correlation between subjective sleepiness and alpha burst activity was also significantly correlated with total work hours. Therefore, while measurements of sleepiness may differ, researchers generally accept that if a person subjectively states they are tired, they probably are. Additional research is needed to clarify the circumstances under which some measures are more useful than others.

Three areas that have been investigated extensively have to do with clinical sleep disturbances, circadian rhythms, the effects of fatigue on performance and alertness, and the ways of measuring alertness.

Clinical Sleep Disturbances
Fatigue and alertness can also be affected by the presence of various biological and physiological conditions that affect sleep and wakefulness. The disturbances in sleep and wakefulness associated with working irregular hours are not considered biological or clinical disturbances of sleep. These pathological disturbances in sleep and wakefulness must meet various diagnostic criteria in order to qualify as clinical conditions.

There are several different types of sleep disturbance: insomnia, which refers to too little sleep; hypersomnia which refers to too much sleep; and parasomnia which refers to deviation from normal sleep patterns. Some authorities estimate that about one third of the population suffers from disturbed sleep (Bixler, Kales, Soldatos, Kales, & Healy, 1976; Liljenberg, Almquist, Hetta, Roos, Agren, 1988)

According to these specialists, in order to meet the criteria of chronically disturbed sleep a person must suffer some form of sleep disturbance at least every other day for a period of three weeks. These disturbances must involve either taking to much time to fall asleep (more than 45 minutes), repeated waking (more than 5 times per night), waking up too early (more than 60 minutes) or getting one and a half hours less than the required 8 hours of sleep. Moreover, any of these symptoms must also be accompanied by disturbances in performance or daytime functioning.

Insomnia is present in about 5 to 6 percent population. In most cases insomnia is thought to be the result of learned psychological responses. Persons suffering from insomnia often report high levels of anxiety associated with worries, traumatic events, or prolonged stress from work or other sources. Depression is another common source of disturbed sleep. Depressed patients often report obtaining less sleep, have difficulty falling asleep, intermittent awakening, and early morning waking. Treatment usually involves some form of psychological intervention including cognitive restructuring and relaxation training.

Hypersomnia, the desire for more sleep, usually manifests itself as a difficulty in staying awake. Criteria for diagnosing this condition are a consistent inability to remain awake in typical everyday situations such as traveling as a passenger in a car, watching TV, listening to a lecture, or reading a newspaper. Common causes of hypersomnia that have received increased attention recently are snoring and sleep apnea which both have excessive sleepiness as associated symptoms.

Approximately 30 percent of men and 20 percent of women experience snoring. Recent studies have demonstrated that snoring reduces both the quality and duration of sleep as a result of short periods of waking and the condition has also been connected to the occurrence of high blood pressure. It is hypothesized that sleep apnea is the result of temporary blockage of the respiratory pathway due to excess fatty tissue in the throat or a relaxation muscles of the throat. When the person has difficulty breathing a startle response occurs causing the person to momentarily wake up to get air. Interestingly, the person is unaware of the recurrent awakening that occurs throughout the sleep period. Unfortunately, the repeated awakenings reduce deep sleep and REM sleep. REM sleep (Rapid Eye Movement) is a recurring portion of a normal sleep pattern associated with dreaming and believed to restore brain function. A person suffering from sleep apnea experiences extreme tiredness during the day. Recent research has begun to look at the relationship between sleep apnea and involvement in accidents.

There are several treatments for sleep apnea and excessive snoring. Sleeping on ones side can reduce snoring and sleep apnea. Surgery can be used to remove obstructions. Weight loss may aid in decreasing the amount of fatty tissue in the throat. Another treatment is the wearing of an appliance in the mouth to keep the airway open. Known as CPAP, continuous airway pressure, the procedure has been shown to be successful.

Parasominas are disturbances during sleep, which disrupt but do not prevent sleep. The most common are nightmares, sleep walking, and bruxism or gnashing of teeth.

While these disorders are not the result of working in a railroad environment the co-occurrence of these disorders in conjunction with an irregular working schedule found in railroad settings may perhaps lead to increased risk for performance decrements. Consequently, railroad companies have sought to engage in identification of persons with sleep disorders to minimize the likelihood of problems arising.

Circadian Rhythms
There have been many studies that have demonstrated that there is a natural pattern of sleep and wakefulness in many organisms. In 1991 the Office of Technology Assessment released a report (U.S. Congress, 1991) in which the implications of circadian rhythms for shift workers were examined. Circadian rhythms are thought to be generated by an internal biological mechanism located in the region of the brain known as the suprachiasmatic nucleus. Circadian rhythms persist even in the absence of information about time of day or night. When circadian rhythms are disrupted human function can be affected, performance may be impaired, and a general feeling of debility may occur until realignment is achieved. Shift work may affect circadian rhythm and its effect on performance and sleep-wake cycles are major concerns. The peaks of performance have been found to vary according to the peak in the circadian rhythm of body temperature. Thus, sleep deprivation and variation in the circadian rhythm can affect performance. (U.S. Congress, 1991)

Other research has shown that people vary in their ability to adjust to shift work, with some individuals suffering few if any problems and others finding such conditions intolerable. The OTA report concluded that "An inappropriately phased circadian system ... can result in feelings of malaise and fatigue, disrupted sleep, and attempts to perform certain tasks at a less than optimal time in the circadian cycle. The degree to which individuals are affected by circadian disruption varies, with some people being better able to tolerate circadian desynchronization than others". (p. 15) Tasks involving signal detection, reaction time, and handling of simple arithmetic correlate with circadian changes in body temperature, peaking during the afternoon, while performance of cognitive tasks involving memory may peak in the morning (Colligan, 1983; Colligan & Tepas, 1986)

While relatively few studies have recorded 24-hour real task data in the field the data that do exist document decreased performance at night. Studies have shown that speed at which a task is performed decreases at night (Browne, 1949; Wojtczak-Jaroszowa, et. al. 1967), the probability of an error, missing a warning signal or nodding off while driving is highest at night (Folkard, Monk, and Lobban, 1978; Hildebrandt, Rohmert, & Rutenfranz, 1974). Several studies have found that the risk of accidents involving truck drivers between midnight and 4 am to be more than double the average during the day (Jones & Stein, 1987; Hamelin, 1987). The GAO (GAO, 1992) study cited above found that more human factor caused rail accidents occurred from 2 a.m. to 6 a.m. than in any other 4-hour segment and that the accident rate between 2 a.m. and 6 a.m. was higher than at other times.

A recent study of 80 locomotive engineers examined the amount of sleep and time of sleep onset (Reid, Roach, & Dawson, 1997). The results show that the timing of sleep and sleep onset follows a "distinct circadian rhythm, with proportionally more (engineers) sleeping at night." In addition, Baker, Reid, Roach, and Dawson (1997) found that engineers going off duty between 11 PM and 3 AM with 12 hours off got about 6 hours of sleep. Much longer time off periods were required to produce sleep that lasted 8 hours. Thus, despite their irregular work schedules, locomotive engineers sleep patterns follow a circadian rhythm.

Rest breaks and napping are fatigue countermeasures that have received some study. The evidence that naps improve performance appears to be more consistent than the evidence that rest breaks improve performance. According to Luna (1997), "there is a relatively large body of literature documenting the beneficial effects of naps taken during and prophylactically before, continuous operations." For example, Harma (1989) showed that persons taking prophylactic naps were more alert on the night-shift than those that did not. Worksite napping strategies are used in some international air traffic control operations (Costa, 1993). Naps have been shown to be effective in improving performance and reducing subjective sleepiness, especially if they are taken before a prolonged work period. Dinges (1994) noted the effects of "prophylactic" napping and also found that naps of only "25 minutes" could improve performance for hours afterwards.

Prophylactic naps should be distinguished from replacement or maintenance naps. Prophylactic naps are taken prior to a period of sleep loss (e.g. prior to an initial night shift). Replacement naps are taken after a period of sleep loss. Maintenance naps are a form of replacement naps that occur on duty. Rosekind, Graeber, Dinges et al (1994) demonstrated the beneficial effects of strategic on-duty napping with airline flight crews in maintaining alertness and performance.

Dr. Moore-Ede in The 24 Hour Society suggests that a 10-15 minute nap is "ideal" for sleepy drivers. He recommends avoidance of longer naps which "can leave you worse off than before." Ferrer et al (1995), on the other hand, recommends either a 30-minute or 3-hour nap as the "best times for short naps."

Sleep Inertia
Sleep inertia or grogginess after awakening is "ubiquitous" and occurs after awakening from sleep at anytime of the 24-hour day or after a nap. (Dinges, 1989). In some studies naps have been followed by a period of "sleep inertia" immediately after awakening. Sleep inertia can be brief lasting from 1-5 minutes in non-sleep deprived subjects. However, sleep inertia can last longer if the nap follows a prolonged period of wakefulness. The duration of these effects however, is usually "so brief that they are ignored in many investigations of the effects of naps in adults." (Dinges, 1989, p. 195) Never the less, "sleep inertia can be severe if the nap is taken by someone with a severe pre-nap sleep debt." (p. 50) The degree of sleep inertia reported appears to be correlated with the time of day that the nap is taken and the length of time since the last major sleep period. Interestingly, reaction time performance was directly related to sleep stage at awakening. Persons awakened during stage 4 sleep (the deepest) yielded maximum reaction times (Akerstadt, Torsvall, and Gillberg, 1989).

This grogginess can have a negative influence on performance lasting as long as 20 to 30 minutes (Taub et al, 1977; Taub, 1979). Dinges (1995) reported that the advantages of napping are not realized until the negative effects of sleep inertia are overcome. However, sleep inertia can usually be reversed within 15 minutes by activity and noise. Some studies suggest that sleep inertia may last anywhere from 5 minutes to up to one hour after awakening. Other researchers suggest that sleep inertia typically lasts 15 to 30 minutes after awakening. One researcher recommended that napping aircrews be monitored for 10 minutes after awakening until sleep inertia passes (Ferrer et al, 1995). It should be noted that most reports of the effects of sleep inertia are found in sleep deprivation studies in which study participants had been awake for long periods of time. Rosekind, et al (1996) suggests limiting naps that are immediately prior to work to 45 minutes in order to minimize the chance of entering deeper sleep.

Deep, slow wave sleep has also been associated with sleep inertia. The degree of sleep inertia appears to be correlated with the timing of the nap, the depth of sleep and the length of prior wakefulness. Sleep inertia has been shown to increase when associated with naps after long periods of prior wakefulness, naps taken during the first few hours of nocturnal sleep or during the circadian trough, and with deeper (slow wave) sleep.

Not withstanding these negative effects, Dinges (1985) indicates that sleep inertia can be anticipated and planned for if a worker must wake quickly and respond to immediate performance expectations. More importantly, Rosekind (1997) noted that in an emergency situation the effects of adrenaline can quickly overcome the negative effects of sleep inertia.

Napping has also been associated with other negative effects. The Office of Technology Assessment’s report on Biological Rhythms and Work Schedules cautions shift workers to avoid off-duty naps if the next major sleep period is at night. Off-duty napping may interfere with a permanent night worker’s ability to sleep normally. Rosa et al (1990) hypothesized that napping may hinder adaptation to shift work by "providing an excuse for sacrificing regular sleep, by making a person too drowsy upon awakening, or by slowing the inversion of the circadian rhythm." The latter concern may be less important for rotating shift workers or permanent shift workers who often do not maintain the same sleep schedule on days off as on workdays. Torii et al. (1982) in a study of nurses on occasional night shifts found that naps taken before and after night work reduced the total sleep time during the subsequent night’s sleep. Long-term effects of naps on performance, alertness and adaptation to shift work have not been defined. One study (Bonnet and Arand, 1994) indicated that individuals who had several one-hour naps during night work experienced larger performance decrements than those who had a prophylactic four-hour nap and remained awake all night.

Napping and Performance
According to Dinges (1989) there have been more than a dozen studies investigating the effects of napping on mood, performance and psychophysiological activity. Mood variables and self-reported alertness have improved in most studies. One study observed a statistically significant improvement in performance in the hour after a nap. (Taub, et. al 1976). Other studies did not find such an effect. However, all of these studies involved subjects who had taken either a morning or an afternoon nap following a normal night’s sleep. Several studies reviewed by Dinges (1989) reported improved performance several hours after the nap and can have a positive effect as much as 10 hours after a nap. Helmus, Rosenthal, Bishop, Roehrs, Syron, and Roth (1997) reported that sleep deprived normal volunteers were more alert after a 120-minute nap than were narcoleptic patients similarly sleep deprived.

A recently published study by Gillberg, Kecklund, Axelson, and Akerstedt (1996) investigated the effects of a 30-minute daytime nap on alertness. After one normal night’s sleep participants received only 4 hours sleep and were tested every hour thereafter. After 10 hours awake EEG sleepiness and subjective sleepiness increased while vigilance performance decreased. A short nap (mean 19.8 minutes, standard error 2.8 minutes) brought performance back to baseline levels.

Another recent study by Reyner & Horne (1997) found that a short nap of less than 15 minutes (mean =12.4 minutes, standard error =1.2 min) combined with a cup of coffee containing 200 mg of caffeine reduced the number of traffic "incidents" in a driving simulator "3-4 fold." There was no evidence of "sleep inertia" reported. This may have been due to the lack of an accumulated sleep debt.

Most studies of short on-duty naps may not be directly applicable to operational environments and to shift work. An exception is the NASA Ames Research Center study on planned cockpit napping. (Rosekind et, al, (1994). The study demonstrated that on-duty naps, averaging 25 minutes in length, improved performance and alertness in aircrews on long-haul flights. The Ames study appears to be well designed and does support on-duty napping as a promising intervention in a controlled operational setting. Dinges (1995) commented however, that the Rosekind (1994) report demonstrated five "fundamentally important points about using planned napping as a fatigue countermeasure strategy":

  1. It was possible to safely and effectively plan ahead of time for when a nap would be taken
  2. It was possible for every operator to fall asleep in a reasonable period of time
  3. Sleep inertia did not pose a serious problem because 20 minutes was allowed for its dissipation prior to assuming duties
  4. As in laboratory studies the nap improved objective measures of alertness … but did not eliminate feelings of fatigue
  5. The beneficial effects of a single nap were most evident on night flights, when control crews showed increasing fatigue relative to crews allowed a nap. (p. 51)

In conclusion, several summary points about napping can be made:

bullet A "nap" may be defined as any sleep that is less than 50% of an individual’s average nocturnal sleep length (Dinges et al, 1989)
bullet Napping can be either voluntary or involuntary
bullet Planned napping may be an effective fatigue countermeasure in certain controlled settings (Rosekind et al , 1994)
bullet Sleep loss increases the likelihood of napping at any time
bullet The napping environment should be conducive to sleep
bullet Napping can be followed by sleep inertia depending upon how fatigued/sleepy the person is prior to a nap
bullet Many studies show that naps can improve alertness and performance under specific controlled conditions

In general, the effects of napping, following the elimination of sleep inertia, have positive effects on performance that can be seen as long as 10 hours after a nap has been taken. However, it appears that napping research that utilizes a methodology readily generalizable to the on-duty activity of railroad employees is scarce. Thus, definitive conclusions about the effects of napping on the actual day-to-day performance for railroad employees are premature. Further study of the duration and timing of naps in the work/rest cycle, and sleep inertia is needed to clarify the best utilization of this technique.

Sleep and Performance
The effects of sleepiness, sleep loss, and fatigue have been the focus of literally hundreds of studies dating back to a study conducted by Patrick and Gilbert (1896) at the University of Iowa. These scientists studied the effects of keeping a group of subjects awake for over 90 hours. Using performance tests measuring reaction time, motor speed, and memory they demonstrated the deleterious effects of sustained wakefulness.

The "modern" study of sleep and performance began with the work of Williams (1959) who demonstrated that there was a progressive increase in reaction time across days of sleep deprivation. These findings wer evident regardless of the nature of the reaction time task, the duration of the task, and whether the person received feedback on how they were doing.

A little over 100 years after Patrick and Gilbert (1896), Heslegrave and Rhodes (1997) reported that air traffic controllers perceived some degree of performance impairment the end of an 8-hour day and evening shift, but significantly greater performance impairment at the end of an 8-hour midnight shift. In fact, the degree of performance impairment at the end of an 8-hour midnight shift was similar to that of the end of a 12-hour day or evening shift. In terms of sleep, when air traffic controllers (ATCs) worked the midnight shift, they reported only about 5 hours of sleep on a daily basis and only about 6-6.5 hours on day shifts. In terms of this aging population, controllers reported more difficulty with shift-work beginning between 35-39 years of age.

In addition to perceived performance decrements, performance changes across various shifts were also demonstrated. For example, using timed reaction time, reasoning, and spatial relations tasks, ATCs performance began to deteriorate 5-10% on the second midnight shift and by the fourth midnight shift a reduction in performance of 10-18% from baseline was observed. For the EDDMM (evening-day-day-midnight-midnight) shift, significant performance deterioration did not occur until the midnight shifts with a 6-12% reduction in reasoning,. Spatial orientation, and pattern recognition. For the EEDDMM shift, performance impairment of 5-15% was evident during the second day shift and during the midnight shift.

Luna, French, and Mitcha (1997) also reported that ATCs on the night-shift of a forward rapid rotation shift schedule appeared to be falling asleep and reported increasing confusion and fatigue.

Many railroad employees do not have a predictable work schedule like the ones described above. One study by Hildebrandt, Rohmert, and Rutenfranz (1974) found that locomotive engineers were more likely to have difficulty using their safety alerters during the night. A laboratory study by Thomas, Raslear, and Kuehn (1996) also reported some indications of performance decrements in locomotive engineers. Never the less, the implication from these studies is that employees with an erratic work schedule (i.e. one that doesn’t correspond to the so-called typical 8-hour day) may be more likely to experience performance deterioration and decrements.

Summary of Effects of Fatigue on Performance
A number of key findings have been summarized by various authors over the past few years relating to fatigue and rest in the occupational setting involving continuous operations. The following list was generated from a review of several articles published by Rosekind (1995, 1996) and Dinges (1991, 1995). They are listed in the form of points to aid in grasping the significance of the findings. Those interested in a more detailed review of the findings should consult the original sources.

bullet Sleep deprivation results in cognitive performance deficits.
bullet Disruption of circadian rhythms leads to a decrease in performance.
bullet Human beings are not very good at estimating their current level of alertness.
bullet Repeated disruption of sleep schedules can lead to decreased performance.
bullet Sleep inertia can lead to performance decrements
bullet Inability to get regular sleep may lead to disruption of the circadian rhythm.
bullet Short naps have been found to restore an individual’s capacity for performance under certain conditions.
bullet Time off alone may not guarantee a rested workforce. Education, planning, and predictability are needed to maximize utilization of work/rest schedules.
bullet With repeated loss of sleep a sleep debt builds up over time
bullet With increased sleep loss and increased sleepiness a person may become vulnerable to performance problems
bullet Quality of sleep is an important factor. Poor quality sleep can leave a person feeling fatigued and non-restored
bullet There can be a discrepancy between how people are feeling and how sleepy they are physiologically.
bullet Scientific evidence suggests that being on an altered shift schedule, like nights, does not lead to an altered internal circadian pattern.
bullet Shift workers that go back and forth between shifts experience more difficulties between the circadian rhythms and sleep times.
bullet Moving a shift schedule forward involves easier physiological adaptations

The following conclusions regarding the effects of sleep loss and sleepiness were abstracted from several articles written by David Dinges of the University of Pennsylvania (Dinges, 1991, 1995). They cover several aspects of research, which are relevant to transportation operations.

1. Fatigue affects performance

bullet Fatigue erodes performance such as vigilance, sustained attention,
bullet Four main processes affect performance decrements: circadian phase, acute sleep loss, cumulative sleep loss, and sleep inertia.
bullet Fatigue can cause lapses or micro sleeps
bullet Night work which interferes with the circadian rhythm can cause lapses to increase 4 to 10 fold
bullet The longer one is awake beyond 14-16 hours the greater the occurrence of lapses.
bullet Chronic under-sleeping creates a cumulative sleep debt, which can produce a cumulative increase in lapses on vigilance tasks.
bullet Sleep inertia creates lapses.

2. Fatigue produces a variety of performance decrements:

bullet Performance variability
bullet Slowed physical and mental reaction time
bullet Increase in number of work related errors
bullet Increased tendency to persistently repeat behaviors
bullet Increase in false responding
bullet Increases memory errors
bullet Decreased vigilance
bullet Reduced motivation and laxity

3. Magnitude of fatigue effects vary by individual

This limited review should serve to alert readers that the "science of fatigue and alertness" is still in its infancy. Those looking for definitive "proof" of a particular point of view are likely to be disappointed. The results of investigations to this point can only be used to generate basic guidelines to follow in developing fatigue countermeasures.

Summary of Effects of Fatigue on Performance
Source: School of Education - University of Denver

A number of key findings have been summarized by various authors over the past few years relating to fatigue and rest in the occupational setting involving continuous operations. The following list was generated from a review of several articles published by Rosekind (1995, 1996) and Dinges (1991, 1995). They are listed in the form of points to aid in grasping the significance of the findings. Those interested in a more detailed review of the findings should consult the original sources.

bullet Sleep deprivation results in cognitive performance deficits.
bullet Disruption of circadian rhythms leads to a decrease in performance.
bullet Human beings are not very good at estimating their current level of alertness.
bullet Repeated disruption of sleep schedules can lead to decreased performance.
bullet Sleep inertia can lead to performance decrements.
bullet Inability to get regular sleep may lead to disruption of the circadian rhythm.
bullet Short naps have been found to restore an individual’s capacity for performance under certain conditions.
bullet Time off alone may not guarantee a rested workforce. Education, planning, and predictability are needed to maximize utilization of work/rest schedules.
bullet With repeated loss of sleep a sleep debt builds up over time.
bullet With increased sleep loss and increased sleepiness a person may become vulnerable to performance problems.
bullet Quality of sleep is an important factor. Poor quality sleep can leave a person feeling fatigued and non-restored.
bullet There can be a discrepancy between how people are feeling and how sleepy they are physiologically.
bullet Scientific evidence suggests that being on an altered shift schedule, like nights, does not lead to an altered internal circadian pattern.
bullet Shift workers that go back and forth between shifts experience more difficulties between the circadian rhythms and sleep times.
bullet Moving a shift schedule forward involves easier physiological adaptations

The following conclusions regarding the effects of sleep loss and sleepiness were abstracted from several articles written by David Dinges of the University of Pennsylvania (Dinges, 1991, 1995). They cover several aspects of research, which are relevant to transportation operations.

1. Fatigue affects performance

bullet Fatigue erodes performance such as vigilance, sustained attention,
bullet Four main processes affect performance decrements: circadian phase, acute sleep loss, cumulative sleep loss, and sleep inertia.
bullet Fatigue can cause lapses or micro sleeps.
bullet Night work which interferes with the circadian rhythm can cause lapses to increase 4 to 10 fold.
bullet The longer one is awake beyond 14-16 hours the greater the occurrence of lapses.
bullet Chronic under-sleeping creates a cumulative sleep debt, which can produce a cumulative increase in lapses on vigilance tasks.
bullet Sleep inertia creates lapses.

2. Fatigue produces a variety of performance decrements:

bullet Performance variability
bullet Slowed physical and mental reaction time
bullet Increase in number of work related errors
bullet Increased tendency to persistently repeat behaviors
bullet Increase in false responding
bullet Increases memory errors
bullet Decreased vigilance
bullet Reduced motivation and laxity

3. Magnitude of fatigue effects vary by individual

This limited review should serve to alert readers that the "science of fatigue and alertness" is still in its infancy. Those looking for definitive "proof" of a particular point of view are likely to be disappointed. The results of investigations to this point can only be used to generate basic guidelines to follow in developing fatigue countermeasures.

Technological Countermeasures
Source: School of Education - University of Denver

The NHTSA Office of Crash Avoidance Research has funded the University of Pennsylvania to conduct laboratory experiments to evaluate the validity, sensitivity, and reliability of selected personal fatigue detection devices and measures, including eye closure measures such as PERCLOS, a measure of eyelid droop. Other psychophysiological measures assessed included two eye blink measures (e.g. MOMs), two electroencephalograph (EEG) measures, and a head movement detector. All measures had some validity, but the results most strongly supported the validity of PERCLOS. FHWA and NHTSA believe that PERCOS is the "most promising real-time" measure of driver alertness for in-vehicle systems. Publication of these findings is in process.

Countermeasures Strategies
There are many factors likely to affect the level of fatigue and alertness that an employee will experience. An employer cannot control all of these factors. An individual cannot control all of these factors. However, given the current state of research and knowledge, the following points have been summarized from the proceedings of the Fatigue Symposium sponsored by the National Transportation Board and NASA Ames Research Center (1995), several articles by Rosekind (1994) and Dinges (1995). They may be helpful in identifying components of industry pilot projects that are useful in alleviating the performance effects of fatigue in the railroad industry.

bullet Educating employees and their families on the effects of fatigue and ways to increase rest may decrease the likelihood of fatigue
bullet Education related to enhancing the factors affecting the quality of sleep can affect fatigue
bullet Maximizing the likelihood of a person getting 8 hours of uninterrupted sleep may contribute to a decreased likelihood of fatigue
bullet Decreasing the likelihood that a person’s circadian rhythms are disrupted may decrease fatigue
bullet Under certain controlled conditions short, planned naps may restore a fatigued individual’s capacity for performance
bullet Minimizing repeated sleep loss may decrease the likelihood of the development of a sleep debt which can negatively affect performance

Past and Current Developments
Source: School of Education - University of Denver

Recent Developments in 1999
As the railroad industry's efforts to address employee fatigue have progressed, various companies have begun to develop formalized fatigue mitigation programs. Each of the major railroads has continued its efforts in this regard through their participation in NARAP, the AAR Work Rest Task Force, the National Carriers Council, and by sponsoring various pilot projects and programs.

On a national level railroads and labour signed a national agreement in principle that addressed several fatigue issues. One of the most important aspects of this agreement was the understanding that future fatigue programs would essentially be "cost neutral" in overall impact. This agreement enables both management and labour to avoid many difficult and contentious contract issues.

Several projects have been undertaken by various railroads that have continued to explore ways to successfully implement fatigue management in the industry.

Amtrak Current Developments
In 1999, Amtrak joined the other major railroads in beginning a fatigue management program. Utilizing the services of an outside consultant the railroad began a multiphase project designed to address the fatigue issues faced by its employees. In Phase I a series of educational seminars was held to begin improving awareness of fatigue issues. Six such seminars were held in locations across the United States including:including Denver, San Antonio, Chicago, Jacksonville, and Richmond. Participants included conductors, engineers, management, union leaders, and FRA staff. In addition, a survey and interview of 142 employees were also conducted.

In phase II of the project a Fatigue Risk Assessment was conducted on all crews in the Jacksonville to Lakeland (Florida) area. This included interviews and assessments designed to determine the extent to which employees were at risk to be in the Circadian Alertness Scale that would be considered fatigued. These results were discussed with Amtrak officials in July of 1999. Based on these discussions a pilot region was identified that would be used to determine the overall level of fatigue that the individual might experience.

In phase III a steering committee was formed to oversee the pilot projects. Next, a series of local task teams were formed to implement the program. The task teams then met with individual crews that worked the Jacksonville to Lakeland run. This segment of the phase was useful in determining employees needs and perceptions. This information would then be used in determining possible interventions. Finally, in phase III measurement tools to determine the effectiveness of the program were identified.

The measurement tools used in this project will be very similar to those used in the Conrail IMPAC and Canadian Pacific CANALERT projects. A multi-method approach to measuring fatigue will be implemented with measurements being obtained in the following areas: self-report measures, scientific questionnaires, sleep/rest/alertness logs, activity monitors, and Circadian Alertness Scaling (CAS). The CAS is a model used to approximate laboratory data that indicates sleep and alertness. Prior to implementing any significant interventions, baseline data has been gathered that will be used to evaluate the effectiveness of interventions.

Currently, measures to reduce fatigue that are under consideration include fatigue management training, napping policies, rest facility standards, sleep apnea screening, alertness monitoring, crew scheduling, and the implementation of various policies and procedures designed to affect operational practices to improve fatigue. Completion of the pilot project is not expected for at least six to twelve months at which time additional measurements will be completed to determine the effectiveness of the program.

BNSF Current Developments
Work/Rest Scheduling Projects
Thayer Project

During June 1998 BNSF undertook a major project to upgrade and maintain the track, bridges, and other elements of the Thayer, Missouri Sub-Division. In order to complete the massive project more than 700 BNSF employees plus additional contractors were employed. This project was designed to employ as many as 900 railroad employees in a 12-day blitz. During this period the crews resurfaced about 115 miles of track, rebuilt 177 grade crossings, installed 46,000 wood and 41,000 concrete ties, resurfaced 115 miles of track and rebuilt 8 bridges. In order to accomplish the production goals for this particular sub-division it was planned that crews would work around the clock. During this time all regular traffic was re-routed to alter native routes.

Prior to beginning the project, a, a plan was developed to ensure that fatigue issues and concerns were considered were considered when developing start-times and shift cycles. A detailed briefing was prepared for each work group addressing issues related to start time, amount of rest needed, and additional information on obtaining rest. These briefings were prepared and distributed to various workgroups and local safety personnel.

As planning for the project progressed it became apparent that there would be an opportunity to investigate some assumptions about the effects of various start times on fatigue. Accordingly it was decided to prepare questionnaires and utilize actigraphs on several key work groups thought to be representative of other employee groups throughout the project.

The two groups that were extensively studied were the crew of the P8-11, consisting of about 100 employees and a crew rebuilding a bridge at milepost 442, consisting of about 55 employees. The P8-11 is a specially constructed set of machines that work in sequence preparing and replacing roadbed, ballast, ties, and track. Crewmembers must operate the machines and keep them in proper sequence, at a steady rate, in cooperation with other equipment.

The P8-11 work group began the project with a 4 :00 A.M. start time. Questionnaires were administered to the personnel at the end of the shift.

Figure 10. Thayer Average Start Time by Day of Project.

A need to minimize interference with operations necessitated some flexibility in questionnaire administration times. Preliminary analyses indicate that for 48 questionnaire respondents on the P8-11 crew, the average hours of sleep varied by start time. As can be seen in the above graph, persons starting work at 7:00 A.M. averaged 6.5 hours of sleep. It should be noted that 62.5% of the sample reported getting 6 hours of sleep or less. Put another way, only 37.5% reported getting more than six hours of sleep per night.

Alertness levels for the P8-11 group, on the 7 point Stanford Sleepiness Scale (with 1 being very alert and 7 being almost asleep) showed that the alertness level was an average of 2.94. Self-report stress measured on a 5-point scale where 1= very little stress and 5 indicates a very great amount of stress was 3.1.

Analyses of the 67 questionnaires from the Bridge 442 group showed 5 people receiving less than 6 hours of sleep and 28.4% getting more than 6 hours. This indicates that they obtained an average of 5.6 hours of sleep ranging from 0 to 9 hours of sleep per day. Start times did not vary for this group.

The average alertness level for the group was 3.4 and the average stress level was 2.9. This can be compared to air traffic controllers working the night shift averaging 2.5; BNSF pool crews without assigned days off averaging 3.4 and BNSF extra board train crews without assigned days off averaging 3.6.

Figure 11. Thayer Stanford Sleepiness Scale Scores

The accomplishments of this project are more remarkable for the impact that they had on the various groups than the overall intervention. Specifically, the effect of studying the fatigue levels of the employees brought to the fore the issues of managing fatigue. During the study supervisors kept changing the work schedule as they became more aware of the fatigue issues

While the information developed during this project lead to some immediate interventions, more importantly, the effect of studying the fatigue levels brought to the forefront the issue of managing fatigue. During the study supervisors kept changing the work schedule as they became more aware of the fatigue issues. For example, the P8-11 crew initially had a start time of 4:00 A.M. and it was expected that crews would be working about 12 hours per day. As the project got underway data began to show that many people were working 14 hours or more a day with some working as much as 18, or 20 hours. Immediate action was taken to divide the work group into 2 overlapping groups to reduce the time at work and thereby increase the potential for hours of sleep.

According to anecdotal reports provided by persons gathering the data, BNSF management on-site noticed that the fatigue levels of the employees seemed to be growing. They subsequently allowed them to "sleep in" one day to recover some sleep. In addition, near the end of the project the supervisors in charge determined it would be possible to reduce the length of the workday while maintaining the required level of production. They responded by not only reducing the hours worked but also changing the start times so there was a positive impact on alertness. Overall, the project signals the start of efforts to address fatigue issues in the non-operating crafts.

Bakersfield Project
In the Fall of 1998,. the the AAR and the BNSF agreed to investigate the latest attempt to develop a scheduling agreement for locomotive crews consisting of 10 days on and 5 days off for locomotive train crews. A scientific evaluation of this scheduling arrangement would provide an object measure of its effectiveness. The project was undertaken at Bakersfield, California due to the interest of local labor leaders and the willingness of management and labor to work together to develop a mutually agreeable plan. Several methods were used in the assessment including self-report, objective, and performance based measures of fatigue.

Crews in this study operate from Bakersfield to Barstow, CA, a distance of approximately 140 miles. The mostly single-track territory consists of 49 miles of mountain operation (owned and dispatched by Union Pacific Railroad whose own trains also operate on the tracks) with the balance having an undulating profile across the desert. Traffic is heavy and consists of a variety of intermodal, unit and manifest freight trains. For BNSF, traffic is predominately westward in the early part of the week and eastward later in the week with approximately 8-10 movements daily. Manned helper locomotives and distributed power are commonly used in the grade territory. Some helpers turn back to Bakersfield at the summit, others operate through to Barstow.

The study involved the pre and post assessment of both the engineers who were involved in a 10 - 5 schedule. A baseline observation period was conducted during October 1998 and post testing was conducted in December 1999. The results of these assessments are reported below.

Self Report Measures
The measurement of fatigue through the use of self-report measures involves the detection of the subjective feeling of sleepiness, alertness, exhaustion, and a number of mood-related psychological states. The project at Bakersfield employed several different assessment techniques utilizing self-report measures. As can be seen from the graph displayed in the chart below there was a significant difference between the scores at the pre testing phase and the post-testing phase for participants in the study. The lower the score the greater the alertness.

Figure 12. Pre-Post Comparison of Alertness Ratings

The next graph indicates that study participants' feelings of exhaustion decreased by the end of the scheduling period in comparisons to ratings taking before the scheduling change took effect. In the chart below lower scores indicate lower frequency of feelings of exhaustion. These ratings also indicated that the engineers were reporting fewer feelings of exhaustion in comparison to the conductors who were not on the 10-5 schedule. Results of the pre post comparison of engineers on the Stanford Sleepiness Scale were non-significant.

Figure 13. Pre-post comparison of Stanford Sleepiness Scores.

Figure 14. Pre-post Comparison of Engineers only on Stanford Sleepiness Scale.

Several items were included in the survey that addressed quality of life issues. For example, one of the areas assessed examined the degree to which respondents reported changes in time for entertainment and recreational activities. As can be seen from the graph below there was a significant increase in the reported degree to which engineers, in comparison to conductors, felt that they had time to spend in entertainment and recreational activities.

Figure 15. Pre-post Comparison of QOL Entertainment Activity.

Additional analyses are included in the following table and indicate that several significant differences were found between the engineers' levels of fatigue when going on duty at pre and post testing.

This measure, along with other similar ones, indicated that this approach has produced a positive change in the lives of the participants in the 10/5 scheduling program. For the most part, participantís comments regarding the effects of the scheduling program were positive. Comments and opinions obtained from labor representatives indicated that approximately 65% of the members of the local union organization were in favor of the program with some modifications. In general, the subjective impressions of this program indicate that it met with moderate approval by the participants and improved or reduced fatigue on several subjective measures.

Other Scheduling Projects
It should be noted that BNSF has several scheduling projects in place at the present time. Superior Wisconsin is still operating on a 10 and 5 work schedule that appears to be running smoothly. There is also a project in Ft. Madison, Iowa that has been running for over a year now. However, BNSF has pulled back from many of its 8 and 3 agreements on extra boards. At this time it is unclear if there are any 8-3 agreements in place. So, with the exception of the Bakersfield project, BNSF has not continued its efforts to implement work/rest scheduling agreements on its property.

Education and Training
Following its initial effort to educate employees as to the nature of the fatigue issues mentioned in the previous section, BNSF has continued to develop additional educational materials on the effects of fatigue and how to manage its impact in the work place. These brochures are distributed to employees as part of the ongoing effort to maintain an awareness of fatigue issues.

In addition, the BNSF has updated the Fatigue Countermeasures Training program based on new information identified by the scientific community. This training module will be presented in mandatory rules training classes for all engineering employees in the second and third quarter of 2000. It will also be presented to all TY&E employees as a module in the Crew Resource Management program that will be introduced in late 2000.

After their initial successes, BNSF decided to step back and evaluate where they have been and where they wanted to go. The implementation of scheduling projects requires a high degree of cooperation and coordination between labor and management. To assist in this process management and labor approached the National Mediation Board (NMB) for assistance. In late 1999 the NMB held training classes for the BLE/BNSF/UTU Work Rest Committee on Interest Based Bargaining. In subsequent meetings the NMB helped that committee establish their future direction. Some of the key initiatives under consideration are 8 hours undisturbed rest, 7:00 A.M. mark-ups, and the expansion of assigned days off program. In addition, BNSF is in the process of expanding their napping policy to include all Engineering employees.

BNSF currently has several initiatives underway to develop assigned days off programs for pool service employees. Many complex operational and work agreement issues need to be addressed. Following the completion of the initial Bakersfield pilot labor and management agreed to hold further discussions on ways to improve the project. Implementation of the second pilot is imminent. It seems that after an initial series of successes, followed by a period of consolidation and evaluation, BNSF stands poised to make further progress.

It appears as if BNSF has decided to regroup from some of its previous efforts in addressing employee fatigue through scheduling initiatives. The Bakersfield project is the only one that appears to be moving forward. Apparently content with its previous education and training efforts BNSF has done little over the past 18 months to address fatigue issues and has apparently taken steps to remove its scheduling initiatives, with the exception of Superior and Ft. Madison. Hopefully, the successes in the Bakersfield area will build momentum for further effort.

Canadian National - Current Status - Canadian National - Current Developments
The Canadian National Railroad has developed an Alertness Assurance Implementation Committee that is comprised of representatives from the BLE, the UTU, Crew Management Strategies, and a Project Officer. In addition, outside consultants from Circadian Technologies, Inc. provide specialized expertise in the design of the scheduled crew work windows. Recently, top management of the CN has reaffirmed its commitment to continue with the Alertness Assurance initiative in the several key areas.

Napping Policy
The CN napping policy was designed to provide quiets area with reclining chairs or couches that could be used by operating crews when delayed prior to boarding their locomotives. As in other companies, crews could use these facilities for short naps to retain or enhance their alertness level during this delay period. In addition, crews arriving at their home terminal could use the nap facility to recover their alertness before driving home.

According to documents provided by management, twenty-nine locations have been targeted for facility upgrade in 2000. Extended run terminal locations were the first priority with secondary main line, single subdivision and operations to follow. A total of 106 chairs have been purchased and distributed to the District for implementation. The complete installation for Western Canada has been targeted for 4th Quarter of 2000.

Rest Facility Improvements
Disturbed sleep is considered a major problem for railway shift workers due to the irregular and unscheduled hours of service. This condition is aggravated when employees are required to use facilities that have poor environmental conditions, peculiar noise levels and light and temperature variance.

CN provides rest facilities for Away from Home operating crews through the use of CN bunkhouses or non-CN commercial hotel/motel establishments. CN reviewed 27 facilities in 1997 for compliance and effected improvements in 17 of them. Improvements required were associated to renovations, blackout curtains, insulation (sound), bunkhouse / motel staff training, and availability of personal noise suppression devices. New facilities at Biggar, Jasper and Boston Bar incorporated design features recommended by CTI.

Educational Activities
In order to provide basic education on fatigue countermeasures CN provided an opportunity for employees to attend a 4-hour workshop while on duty. The training incorporates instruction on the physiological/psychological mechanisms of fatigue, and fatigue countermeasures as well as specific lifestyle training for employees working in a 24-hour environment. Lifestyle training was provided to the Rail Traffic Controllers as well as the Crew Management Center Crew Dispatchers in order to ensure that those involved in the implementation have an understanding of the strategies introduced.

Work Scheduling
Time pools were developed as a means of ensuring predictability and regularity in the lives of employees through the use of calling windows and scheduled days off, using circadian sleep wake principles. Time Pools (schedules) offer significant benefits in reduced absenteeism, improved subjective alertness, improved sleep quality, greater job satisfaction, and increased predictability in railway planning.

Current Schedule Implementation Status

bullet Rivers Sub (Winnipeg to Melville) - implemented June 1999
bullet Edson Sub (Edmonton to Jasper) - implemented July 1999

Schedules have been designed for the Vancouver - Kamloops segment, however, a number of operational changes have delayed initial implementation. It is anticipated that the schedules on this segment will be in place before the end of the first quarter 2000.

In addition to the mainline operations CN is attempting to introduce an enhanced Lifestyle Schedule operation on the Prince George - Smithers segment. Currently under review is the feasibility of scheduling an area with limited traffic. Mainline traffic has a high degree of predictability while the scheduling on the BC North territory presents unique challenges and adjustments. Although all the operational conditions have not been reviewed to date CN is expecting to implement a form of scheduling before the end of the first quarter 2000.

CSX Current Developments in 1999
CSX has continued its program to address employee fatigue including a continuation and expansion of the sleep apnea survey, additional educational efforts, and progress towards the implementation of a scheduling agreement in the Florence service lane.

Sleep Apnea Survey
As discussed in Chapter Two, Obstructive Sleep Apnea (OSAS) is a sleep disorder that consists of episodes of interrupted breathing, resulting in sleep fragmentation and daytime sleepiness. Studies show that OSAS affects 2% of women and 4% of men of middle age (Young, 1993). Researchers believe that occupational factors may play a role in the development of sleep apnea. For example, the variable sleep schedule found among some railroad workers may worsen sleep apnea. Moreover, an elevated presence of OSAS has been document in the trucking industry (Stoohs et al., 1995). Of most concern to the public is the fact that OSAS has been associated with an increased risk of driving accidents (Maycock, 1996) and performance problems at work (Aguirre et al., 1996; Ulfberg et al., 1996). Fortunately, OSAS can usually be treated effectively.

In order to understand the prevalence of OSAS in the railroad industry, CSX undertook a survey of 2386 employees in the Florence Service Lane between August 1998 and September 1999. The project began with an extensive educational campaign addressed to all of the management and union representatives to acquaint them with this sleep disorder. Next, a pre-screening questionnaire was mailed to employeesí homes. Consisting of 37 items, it was designed to determine the presence of sleep apnea. Incidentally, the questionnaire included the Eppworth Sleepiness Scale that has been used extensively in railroad studies of fatigue.

A total of 248 employees, consisting of 240 males and 8 females returned the completed questionnaires, for a response rate of 10.4%. The average age of respondents was 47.3. A significant number of respondents reported lifestyle factors related to sleep apnea including being overweight and smoking. Reported alcohol use was very low. The average score on the Eppworth scale was indicative of moderate sleepiness (Mean = 10.5, sd =5.0). Twenty percent of employees reported hypertension and 61% reported habitual snoring.

Based on a review of the surveys, 136 of the 248 respondents indicated higher risk factors for sleep apnea; 31% indicated hypertension, 41% were overweight and 85% experienced habitual snoring.

All respondents were informed of their results on the screening questionnaire within two weeks. Persons not at risk received a letter stating so along with an explanation of the results. Those employees whose results indicated a high risk for sleep apnea were contacted both by mail and telephone to discuss their results and encouraged to participate in further assessment through their primary care physician and a sleep lab.

The results of the sleep studies are reported in the following table.


Program Step

Number of Employees

Percent of

Total Population

Percent of

At Risk Population

Total Number of Employees 2386    
Questionnaires Returned 248 10.4%  
Identified At Risk 136 5.7%  
Sleep studies performed 56 2.3% 41.2%
Confirmed OSAS 43 1.8% 31.6%

Figure 16. CSX 1999 Sleep Apnea Survey Results

This study suggests that about 2% of railroad employees are suffering from OSAS. This is an extremely conservative estimate since only 10.4% of the population turned in the survey. Many individuals who did not return the survey were probably either not interested or may have already been diagnosed and therefore not needing to complete the survey. It is also possible that employees did not either complete or return the survey out of fear of the potential harm that might arise if their results were not kept confidential. Further studies would have to be done to accurately estimate the incidence of sleep apnea in railroad employees.

The diagnoses of the 56 people who had a sleep study were as follows: 43 had mild/moderate sleep apnea requiring CPAP treatment or surgery; 11 (20%) had mild apnea or primary snoring; one case of narcolepsy; one case of circadian rhythms disorder; several cases of periodic leg movement disorder as a secondary diagnosis.

Many different stakeholders perceived the program as a success. In the final report submitted to CSX management the consultants stated "CSXT employees in other areas, having heard about it (the screening program) through their peers, have called with requests for the Program in their area." (Circadian Technologies Inc., 1999) In all, this program provided a much-needed service to railroad employees.

Napping Policy
As of March 1, 2000 CSX implemented a napping policy that permits road train crews to take naps, under certain controlled conditions. Naps are permitted, when appropriate safeguards have been implemented, for one member of the crew at a time. There must be an alert crewmember observing the equipment and the condition of other equipment or trains as they roll by. At the time of this writing there have not been any problems reported with this policy by operations managers.

Education and Training
A new set of videos has been planned for distribution beginning in May 2000. This set of videos will target the physiological issues and alertness. A special emphasis will be placed on the types of countermeasures that have been found to be effective in the railroad industry. It will include reports of successes at various locations around the CSX system including Russell, Kentucky and the Florence Division Pilot.

Work/Rest Scheduling Project
Florence Service Lane
The FSL project is designed to improve the quality of life and predictability of rest for a 129 mile run between Richmond, Virginia and Rocky Mount, North Carolina. This is primarily a freight pool with runs lasting between 8 and 12 hours.

Working with representatives from the BLE and UTU, a scheduling arrangement of 10 days on, 3 days off was developed. In order to determine whether this approach was effective, an outside consultant was hired to conduct an evaluation of the program. With the participation of Union and Management officials, a study was designed which included several baseline measurements and follow-ups. This design would permit a pre-post within group comparison of the persons in the program. In effect participants are able to serve as their own control group.

Beginning in December 1998, preliminary measurements were obtained on the persons operating in the Florence freight pool. Research assistants greeted employees as they began or ended their assignments and invited them to participate in the study. Those who were interested received a questionnaire and a consent form. Participants wore an Actigraph - Sleep Watch for 14 days, completed a computerized assessment battery for 20 days and provided records of hours worked and hours spent between assignments at home or away-from-home.

Measures used for this investigation are similar to those used in previous investigations. The questionnaire consisted of several different instruments including: the Stanford Sleepiness Scale, Eppworth Sleepiness Scale, Visual Analog Test, Denver Adjective Checklist, and various demographic measures. The computerized assessment battery, the Denver Fatigue Inventory included several measures adapted from the Performance Assessment Battery (Thorn, Genser, Sing, and Hegge, 1985). These included a simple choice reaction time test, the Stanford Sleepiness Scale, the Manikin Test, the Wilkinson Serial Addition and Subtraction Test, and a randomized visual tracking measure. Reliability and validity information on these measures has not been established for the present form of the test. However, previous administrations of the measures have been reported in (Thorn, Genser, Sing, and Hegge, 1985).

Pre-test data on the study participants were gathered at two points prior to the implementation of the intervention of the 10-3 scheduling arrangement. As can be seen in Table 11 the engineers and conductors indicated that they were receiving at least 8.5 and 7.0 hours of sleep on the night before the study was initiated. At the second pre-test period participants indicated that engineers and conductors were receiving 7.5 and 7.15 hours of sleep on the night that the questionnaires were administered. No significant differences were detected between groups.

Similar findings were obtained on most of the other measures. No differences were found between groups on the Stanford Sleepiness Scale, the Visual Analog Scale, the Denver Adjective Checklists, or the Eppworth sleepiness scales.

Table 11. Pre-test data for CSX Florence Service Lane 10-3 project

Pre Test 1


Pre Test 2


N Mean SD N Mean SD sig
Hours of sleep in last 24 hours?
Engineers 11 8.55 3.56 25 7.5 2.39 ns
Conductors 4 7.00 3.83 12 7.15 2.23 ns
Hours of sleep in last 72 hours?                
Engineers 11 21.9 6.90 25 18.63 6.30 ns
Conductors 4 15.67 4.04 13 16.69 5.45 ns
Stanford Sleepiness Scale
Engineers 11 2.73 1.27 25 2.76 1.56 ns
Conductors 4 2.25 1.5 12 2.58 1.83 ns
Denver Adjective Check List
Engineers 11 2.35 0.75 25 2.50 0.99 ns
Conductors 4 2.39 1.02 12 2.23 0.05 Ns
Eppworth Sleepiness Scale
Engineers 11 11.00 4.62 25 9.2 4.19 ns
Conductors 4 11.5 3.87 12 7.67 4.56 ns
Visual Analog Test
Engineers 11 3.70 2.71 25 4.44 3.32 ns
Conductors 4 2.90 1.85 7 4.31 3.67

The schedule changes were agreed upon by the UTU and put into effect near the end of March 2000. Follow-up data will be collected in June. This project is noteworthy due to the many difficulties that were overcome prior to initiating the agreement. In addition to the challenge presented by traffic increases due to the Contrail turnover, the flooding in North Carolina provided additional tests of CSX commitment to the 10-3 program.

One final observation seems to be in order. Looking at the CSX data for fatigue, one is immediately struck by the fact that both locations were reporting less fatigue than other railroad locations prior to the adoption of a scheduling change. While it is impossible to determine from these data alone what is contributing to these numbers, especially given the absence of a suitable control group, it may be that CSX railroad employees are experiencing slightly less fatigue than their colleagues in other locations. Further study is needed to determine the extent of the fatigue problem to identify any serious lack of sleep on a weekly basis, and to measure any performance decrements as a result of inadequate rest in the railroad industry.

Shelby Mine Run "Shifters"
It is important to note that CSX has also widely established a very popular program for mine operations. In various locations in the coal fields of Kentucky and West Virginia, CSX has developed a program of assigning crews to trains. The approach works very well because crews know when they will be going to work and, for practical purposes, when they will be getting off, as the runs to and from the mines are very predictable. Survey data obtained from persons working on the mine run operations indicate that persons are fairly satisfied with the program.

In addition, data suggest that the employees in that region are not experiencing very high levels of fatigue in comparison to other locations. As can be seen from Figure 179 the data suggest that employees in Shelby, a mountainous region in Eastern Kentucky, are lower than the national average on subjective measures of fatigue.

Figure 17. Stanford sleepiness Scores at three CSX locations.

These results are noteworthy for the simple fact that employees coming on duty, in comparison to other locations on CSX and in comparison to a national sample, show lower scores on the Stanford Sleepiness Scale than a national sample of railroad employees. There is not much difference however between Shelby "Mine Run" Shifters and Florence service Lane Pretest study participants coming on duty.

Russell, Kentucky to Columbus, Ohio
Labor and management at Russell, Kentucky have met and implemented an 8-3 work schedule agreement. This program was adopted on January 10, 2000 and has met with widespread acceptance and utilization. Russell to Columbus is on the route that transports coal from Eastern Kentucky to Toledo, Ohio. There are approximately 20 pool turns in this service from April though mid-December when coal is being unloaded in Toledo. The program is successful and is in operation without a guarantee.

UPRR Current Developments in 1999
UP has continued to work towards an effective Fatigue management program for its entire system. The UP has for the most part followed the plan that it developed last year. The company has made efforts to implement almost all of the components of its plan. As mentioned above, AAR Work Rest Task Force agreed on the components of an effective Fatigue Counter-measures Program that was endorsed by the senior railroad executives in February 1998. (Page 32) We will use these eight areas as the basis for commenting on progress in developing fatigue counter-measures. The following lists the most recent developments and accomplishments to date.

Education and Training
In the area of Education, Training, and Communication, UP has developed a number of brochures for employees in the area of fatigue management. This is an improvement over 1998 when only two brochures were available. These brochures address many important issues: 1) Alertness Management Program, 2) Good Sleep Habits, 3) Strategies for Living, 4) Lodging Facilities, 5) Crew Management Systems Wake-íem Up, 6) Planned Nap Program, 7) Crew Scheduling Activities, 8) A Management Perspective, 9) Drugs Alcohol and Fatigue, 10) Fatigue and Family Support. 11) Managers Alertness Travel Guide. 12) Employee Alertness Guide. 13) Brochure and Video Listing Card, 14). Fatigue Myths and Concerns, and 15) Drowsy Driving. In addition, additional information about Non-ops Manager Concerns is reportedly also being developed.

Videos discussing important issues related to fatigue have also been developed. The current list includes such topics such as:

  1. Planned Napping,
  2. Non- Operating Education,
  3. Sleep Basics, and
  4. Alertness Management Program
  5. Alertness Strategies
  6. Family Alertness Awareness and Support
  7. Cost of Fatigue
  8. Sleep Disorders
  9. Womenís Issues and Fatigue
  10. Adolescents and Fatigue
  11. Operating Education

These efforts are an attempt to educate all employees regarding the effects of fatigue and effective countermeasures available to manage it. The UP is making a concerted effort to educate its management and scheduled work force as to the extent of the problem, the need to address the problem. A special edition of The Sleep Solution (Ball and Hough, 1998), a workbook designed to help people get better sleep was distributed to more than 23,000 T&E employees in July 1998 and mailed to all remaining employees (almost 38,000) in November 1998. According to materials supplied by UP management, an additional 325 Sleep Solution books were mailed to new non-operating employees hired in the fourth quarter of 1998 in January of 1999. At the time of this writing approximately 1,044 new Non-Operating and an additional 495 Operating employees had been hired in 1999, all of which received a copy of the book. Newly hired operating employees also get a fatigue component in their new hire training and orientation seminars. In addition, the bi-monthly newsletter On Alert is distributed to all employees (over 60,000) in the UPRR INFO magazine.

Additional alertness management training programs have been developed for new and existing employees in the Harriman Dispatching center, the Employee Assistance Program, UP Police Department, as well as new operating employees, and others. It should be noted that all new employees receive fatigue education as part of their new hire orientation-training program. This includes operations and non-operating employees at all levels of the organization.

Educational seminars continue to be delivered to management and agreement employees as part of the Rules Class Session B Training. All operating employees are required to attend bi-annual rules training. In conjunction with the bi-annual rules training, the employees are also given training in the Alertness Management Program. This program consists of approximately 4 hours of lecture and discussion and video presented by a trained facilitator. A standardized video, Trainers Guide, and Participant Guide were developed specifically for this training program. According to information submitted by the Health Services Department, a significant number of employees have completed the Session B training, which included the fatigue component.

According to information provided by the Health Services Department, more than 70 training programs were given to various groups by their staff in 1999 including shop maintenance employees, contract crew van drivers, and other non-operating managers.

Given the size of the work force and the complexity of the problem the UP efforts are commendable. Considering the complexity of the problem and the nature of relationships between management and labor, the face-to-face training approach is more likely to be an effective method of achieving the educational objectives. UP has made impressive efforts in this area and hopefully, this training effort will continue until all employees have received both the printed and video information and the face-to-face instruction needed.

One development that is particularly noteworthy has been the development of a questionnaire designed to measure the extent to which participants in the educational training seminars have been able to retain the information presented to them. This instrument may be extremely useful in planning continuing and follow-up educational interventions. This program could serve as a model to other railroads that attempt similar training programs. Much time, effort, and resources are expended in the preparation and delivery of alertness and similar programs. In most cases little is done to determine whether these programs actually have the desired effect. If the identified objective is actually worth achieving efforts to measure that achievement are also needed.

Scheduling Projects
The scheduling efforts have been hampered by the fact that complex labor agreements must be negotiated between labor and management. The UP has made considerable progress towards developing an effective collaborative effort with labor in implementing fatigue management and scheduling programs throughout the railroad.

In mid-1997 UP top management developed a set of strategic objectives that would guide the transformation of UPís management culture. One of the strategic objectives was the recognition that UPís employees are critical and imperative to the railroadís success. UP identified and adopted specific strategic initiatives to accomplish those objectives. One of the strategic initiatives specifically address the fatigue and quality of life issues of its train and engine service employees - i.e., to provide more and scheduled time off for UPís train and engine crews. As a part of this initiative, UP commenced developing an effective fatigue management and program. In addition, UPís Health Services Department also began working closely with labor relations to develop an effective strategy in approaching labor with the fatigue management initiatives. As mentioned previously, UP had approached some locations as early as 1996 with programs to implement scheduling changes. While some were met with acceptance, the only agreements still in existence today are the North Little Rock programs. Part of the success of the North Little Rock program was due to the collaboration between management and labor. Management was concerned however, that in attempting to implement new programs around the system that each location would want to negotiate a new contract. The practical limitations inherent in this approach would naturally extend the process over several years. UP approached its labor organizations with proposals to address the fatigue issues through system wide agreements. High level representatives of the UTU accepted this invitation and, at a meeting in Tucson, in February 1999, with management officials from UP, met to discuss their concerns. The meetings established a framework for addressing fatigue issues of train service employees and identified key points and principles that would guide the adoption of crew scheduling agreements throughout the UP system.

These principles were outlined in a document signed April 13, 1999 and included the following provisions:

  1. 8 hrs undisturbed rest when off duty
  2. 7:59 A.M. mark-up after 72 hours or more leave
  3. Not less than 2 days rest after a consecutive number of work days
  4. Guaranteed rest time if work into rest
  5. G-guarantees for employees assigned to the extra board will be prorated on the employees total work cycle days
  6. Significant disincentives will be imposed if persons take uncompensated absences during their assigned work cycle. There was however, no guarantee for those days an employee is on his rest days.

This agreement of understanding developed at the Tucson meeting, provides a concrete foundation of principles and points to which all parties have already agreed. The practical implementation can then be agreed to and put in place, by the UTU local officials, at any location that decides to utilize these work/rest agreements. A similar agreement for freight pool employees was drafted and signed by UTU General Chairmen on July 21, 1999. Also, the first joint UTU/BLE work/rest meeting was held on July 29, 1999.

The events leading up to this agreement and its subsequent adoption in various locations around the UP system are no less than outstanding. By collecting all of the issues related to work rest and fatigue, isolating them into one set of principles, agreeing in principle to focus only on those issues, and to forego negotiations on other perhaps more contentious issues, both the UTU and the UP have laid the foundation for significant progress in this area. One can only be amazed at the exponential progress made in implementing these agreements on over 19 locations throughout the UP system. While other railroads have attempted similar undertakings in the past, hopefully this effort will be sustained successfully over the next few years.

North Little Rock
The two scheduling projects (eleven days on/four days off and twelve days on/four days off) in the North Little Rock area, begun in 1996, are the only scheduling projects that continue from earlier agreements. On-site discussions with employees involved in this project conducted by the author indicate that they are "very satisfied" with the project and do not want to revert to their previous schedule. However, employees also revealed concerns that while the program is a step in the right direction it may do more to improve quality of life than reduce fatigue. This is particularly noticeable when employees are called on-duty just prior to beginning their rest days. A major feature of the project has been that employees are able to identify and plan for days off. However, if scheduling results in working into rest days then both the predictability and rest is compromised. Discussions also revealed that management was satisfied with the agreement and felt that it did not present any obstacles to effectively running the operation. Union officials expressed an interest in adjusting the agreement to include time windows. Management officials have expressed an interest in updating the agreements in that locale to include all of the work/rest principles included in the Tucson Agreement. Such an effort would preserve the existing agreement while improving aspects that will enhance rest.

In early 2000 a questionnaire was mailed to all (N=108) of the UTU employees in the Little Rock area affected by work/rest cycle agreements. The survey developed by the Health Services Department, consisted of 5 questions for employees and 6 questions for families and significant others. Respondents returned 27 surveys (25%) for analysis.

Results of the survey indicated that there was an overall positive perception of the scheduling agreements by both employees and families. Specifically, 67% of employees reported an enhanced quality of life at work while 85% reported enhanced quality of home life. Seventy percent of family members reported that the agreements enhanced quality of life while 45% believed that the agreement had a positive effect.

As previously noted the Little Rock Area agreements have been in place quite a while. These data provide additional support for the positive effects of scheduling projects on employee and family perceptions.

Figure 18. Perceptions of Little rock Scheduling Agreements

La Grande Freight Pool
Following the development and adoption of the principles in the Tucson Understanding, UP began negotiations in earnest with UTU representatives at La Grande Oregon. Subsequently, on June 22, 1999 an agreement was signed between Union Pacific and the United Transportation Union regarding the Work/Rest cycle agreement. This agreement includes the following provisions:

  1. A minimum of 8 hours undisturbed rest
  2. An 8:00 A.M. mark-up after 72 hours off-duty
  3. Work/Rest cycle of 7 days on and 3 days off
  4. An employee who works into their rest cycle may have the rest cycle extended to the full three day rest period or
  5. The employee must take at least 48 hours off
  6. Guarantees will be based on the total number of days worked in a calendar month
  7. No guarantees are awarded for those days a person is on their rest days

La Grande Extra Board
The UTU and the Union Pacific also entered into an agreement for the La Grande extra board on June 22, 1999. The agreement is essentially the same as the agreement for the pool.

Other Locations
Agreements with labor have been obtained at a number of locations on the UP system. In addition, agreements leading to implementation are at various stages of development including implementation, agreement, ratification, or negotiation. According to information provided by UP management in April 2000, 79 locations are at the implementation stage and are using either a 12/4, 11/4, 7/3, or 6/2 work schedules. Another 52 agreements have been ratified and 28 were close to ratification.

While the above information is largely based on management reports that will be verified by on-site discussions with union members by the author, these agreements mark the first time that a work/rest cycle agreement has been put in place on the UPRR since the Alertness Management Program was initiated. Never the less, several key accomplishments should be noted that have implications for the rest of the industry.

First, the fact that it addresses a freight pool is noteworthy. Many of the previous agreements (e.g. 8 and 3) have been made for extra -boards. Also, there, there is a minimum 48-hour rest provision following a consecutive work cycle. This is significant because it adheres to the scientific knowledge that has been gathered regarding the length of time that a person needs to recover from sleep deprivation. In addition, the employee may take the entire amount of time allocated for his/her rest cycle if he/she works into their rest. This was also a significant concern of persons in the North Little Rock area. Finally, this agreement has a work cycle that is only 7 days. Previous agreements were 8-3, 10-5, or 11-4. With regard to the guarantees, the agreements appear to be cost neutral. This is in keeping with the national agreement signed by the National Carrierís Conference Committee and the BLE and UTU. At present, only the UTU has reached agreement on this progressive scheduling approach. Hopefully, it will soon be available to BLE members as well.

Undisturbed Rest
According to materials supplied by management, 100% of the engineers across the system have the option of taking 8, 10 or 12 hours undisturbed rest under their respective collective bargaining agreements. Moreover, employees in the southern region are eligible to voluntarily lay off for two days after working 14 consecutive days. In addition, nearly all of UPís train service employees are governed by collective bargaining agreement rules which permit them to take extra rest ranging in duration from eight hours up to twenty-four hours. The conditions under which such employees make take this rest vary. According to UP Labor Relations officials approximately 85% of UPís employees are covered by such a rule in some form.

Emergency Response Requirements
The Emergency Response Requirements (unplanned work associate with adverse weather conditions or accidents) are currently under discussion. A committee was formed in November 1998 to address these issues. At this point Alertness Solutions Inc., has provided consultation on current Emergency Response policies. In addition, the Safety Assurance and Compliance Program (SACP) Non-Ops group has met and developed a list of issues for both the Engineering and Shop Crafts. However, a final product has not been prepared at the time of this writing.

Planned Nap Program
The napping program has now been implemented system wide for road train and engine employees. Materials supplied by management have noted no safety problems since the program was initiated. This is consistent with the experiences of the Canadian Pacific, BNSF, and Conrail.

An Opportunity Napping Yard Pilot was begun in Houston in October 1999 and will be measured and evaluated for the feasibility of further expansion after 90 days.

Another pilot for mechanical shop employees for Opportunity Napping in Emergency and Unscheduled Work began in October 199 at Hinkle, Houston and North Platte. This pilot will be for 6 months and will then be measured and evaluated.

Considerable activity has been noted in this area. A survey was mailed out in September 1998 to 1,755 employees. Results, based on a 21% return rate, indicated that respondents viewed only 50% of the lodging facilities favorably. Committees have been formed to look into the issues raised by the survey. A special brochure designed to address lodging issues has also been developed for distribution to employees. In addition, a staff person has been assigned to review the issues raised on the survey with the individual facilities.

A second evaluation of lodging facilities was conducted in March 1999. A total of 1,972 surveys were sent out to employees in 3 cities in each of the 3 Regions; Seattle, Oakland, Yuma, Clinton, Jaynesville, Des Moines, Ft. Worth, Dalhart and McAlester. A total of 462 responses were received for a response rate of 23%. Based on the responses to the surveys the Alertness Management Program (AMP) group developed a list of strategies to be used in combating the problems encountered at the lodging facilities. A pilot of this process of responding to the concerns at the lodging facilities was begun in August of 1999.

Figure 19. Comment Card.

Another method, called the Lodging Comment Card, is being piloted beginning in August 1999 for 60 days in the Los Angeles service unit. The lodging comment cards were the result of the 2 separate surveys that were done on the facilities. The previous comment cards simply asked employees to write out their comments. The revised comment card requests specifics on: bedding, room availability, noise, light, etc. and whether the hotel resolved the problem while the employee was there. The revised comment card also includes educational material in the form of strategies and suggestions on the reverse of the comment card itself for the employees to review on good sleep and how to minimize outside disturbances. This comment card can be used as a "mini survey" tool.

Technological Counter-Measures
In previous documents it was reported that UPRR had field-tested two headset communication devices. According to materials supplied by management results of a survey of employees indicated that employees generally do not like or plan to use ear "muffs" or "ear buds" communication devices. At this point the AMP Technology Team has not identified any technologies that are "feasible" at this time.

Status of Union Pacific
Over the past nine years Union Pacific in many ways has led the US railroad industry in addressing the fatigue problem. UP began many studies that documented the importance of fatigue as a health concern for the operating crafts. Beginning in 1997 the UP made a concerted effort to incorporate a thoughtful and directed planning process to systematically address fatigue. UP has provided written materials on the effects of fatigue to its employees. Face to face educational interventions have been delivered to a portion of the workforce. A lodging survey has been administered to about 14% of the workforce, and improvements are underway due to responses that indicated satresponses that indicated satisfaction with only 50% of lodging facilities. Emergency response plans and technological countermeasures have been examined. These efforts address all of the key components of the AAR Work/Rest Task Force and NARAP position statements.

In addition to the above, road train and engine napping programs have been piloted successfully with no reports of operational problems or difficulties. At this point the napping policy is available system wide to UPRR road operating employees.

Most importantly, in the second quarter of 1999 significant progress was made in developing scheduling programs that incorporated scientific principles into the management of fatigue. Employees in many locations will have at least 8 hours of undisturbed rest and the opportunity to request at least 12 hours or more of undisturbed rest. Some areas of the railroad have the opportunity to take longer periods of rest.

At this point, it is safe to say that UPRR has developed a plan that is addressing the major points of the NARAP fatigue Countermeasures plan. While considerable work remains in terms of implementing work/rest scheduling agreements at various locations, and continuing the educational efforts begun, the foundation provided by the Tucson agreement with the UTU will make the eventual adoption of these plans with other unions and the UPRR system considerably less difficult.

Norfolk Southern Current Developments in 1999
Work Rest Committee
In April 1999, the first meeting of the NS Work/Rest committee was held in Norfolk, Virginia. The meeting attendees included representatives of the BLE and UTU along with NS management. The Work/Rest Committee was established to provide a forum to discuss rest and fatigue issues. The Committee met approximately bi-monthly for the last year. The Committee is evaluating work/rest pilot projects as well as identifying other areas for additional pilot projects. According to minutes of the meeting provided the committee is currently reviewing several pilot projects having to do with napping, timely pick-up of road crews, and technological countermeasures.

The committee has addressed a variety of issues at its bi-monthly meetings including: education and training materials and videos; scheduling programs; assigned work/rest days; minimum undisturbed rest; A.M. markups; transportation for crews from trains not arriving at terminals; line-ups; and lodging facilities. These topics are discussed in greater detail below.

Work/Rest Scheduling Projects
Fort Wayne - Chicago
A pilot project began December 7, 1998 and involved 46 pool conductors in the Ft. Wayne - Chicago corridor. Following a 100-day pilot period the UTU ratified the agreement to make it permanent. The notable provisions of this agreement are:

This agreement works with 14 days on followed by two or three days off.

  1. Pool members are split into seven groups with assigned days off.
  2. Assigned days off occur on Monday, Wednesday, or Friday at midnight and allow a minimum of 54 or 78 hours off. These rest days are rotated on a seven-week period to so that each group obtains weekend as well as weekday rest days.
  3. Employees must give forty-eight hours advance notice if they intend to observe their rest days.
  4. Employees cycle off duty at the home terminal at 6:00 P.M. the day before their first rest day.
  5. Employees at the away from home terminal are returned to the home terminal if not called for service by 8:00 P.M. the day before their first rest day.
  6. Pool places are removed from the pool when employees begin their first rest day. Mark up after rest days occurs at 6:00 A.M..

Virginia Division
The Virginia Division project began in July 1999 and involves seven pools. Each pool is divided into six groups with optional rests days available. This project involves both the UTU and the BLE. The notable aspects of the agreement are as follows:

Each group observes a 10-day work cycle and an optional 2-day rest cycle.

Employees cycle off duty at the home terminal at 6:00 P.M. on the day preceding rest. If not called out of the away from home terminal by 10:00 P.M. the employees are deadheaded home in taxis or vans.

Employees who mark off for rest days have their pool turns removed from service. Employees are automatically marked up at 6:00 A.M. on the day following their last rest day.

Based on data provided by NS, this particular project varied in its acceptance and popularity. Reasons for this are not given. In some places rest days were not taken for several months. The project was permanently ratified in two of the districts Roanoke and Norfolk.

Kansas City - Moberly
The Moberly Kansas City corridor is comprised of approximately 35 pool crews on a 131-mile district. The pilot project established a new assigned pool with 8 positions while the number of pool positions in the overflow or unassigned pool become 17 positions. The notable aspects of the agreement are as follows:

  1. An assigned pool of 18 engineers is established to handle the first eight trains out of the home terminal after 4:00 P.M. (2:30 P.M. call time).
  2. Assigned jobs not called within a 24-hour window are paid a basic day for each day missed and stand for the next scheduled trip out of the home terminal.
  3. Assigned jobs are not called to deadhead out of the home terminal.
  4. The overflow pool handles all trains called after the assigned jobs are exhausted and all deadheads out of the home terminal.
  5. Assigned jobs and the overflow pool mark up on a first-in first out basis at the away from home terminal.

The pilot project began on September 28, 1999 and applies only to the BLE. The pilot was scheduled to run for ninety days. No data is available on the current status of the project.

Elkhart - Chicago IMPAC Program
The Elkhart - Chicago was formerly under agreement with Conrail (see previous chapter). Just prior to the acquisition of Conrail the agreement was terminated. It was reinstated, as a pilot project on September 20, 1999. The notable aspects of the agreement are as follows:

  1. Ten hours undisturbed rest at the home terminal and eight hours undisturbed rest at the away from home terminal.
  2. For the first thirty days of the project employees were called at the away-from-home terminal no later than 14.5 hours or deadheaded at 16 hours.
  3. For the next thirty days of the pilot, employees were called at the away-from-home terminal no later than 13.5 hours for service at 15 hours.
  4. For the last sixty days of the pilot, employees were called at the away-from-home terminal no later than 12.5 hours or deadheaded at 14.
  5. Employees who are called and perform service following rest at the away-from-home terminal operate into the home terminal at the completion of their tours of duty.
  6. Each employee is eligible to take four IMPAC (rest days) per month. An employee can elect to have an 11:59 P.M. mark up or automatically receive a 7:00 A.M. markup when returning to service.
  7. No more than 20% of employees in this service will be allowed to be off on Mondays and Tuesdays due to traffic demands and no more than 16% of employees in this service will be allowed to be off during the remainder of the week.
  8. Not more than 50% of the allowable mark-offs will be allocated for IMPAC rest days on each day of the week.
  9. Employees may request up to two additional consecutive days off with at least six hours advance notice.

In November, due to a shortage of engineers, the BLE agreed to reduce the number of engineers off on Monday and Tuesday to correspond with traffic flow. The project is currently still in operation.

The continuation of this project is noteworthy. Not only does it continue one of the earliest established programs that had been put in place, but it speaks to the commitment of the Norfolk and Southern to identifying and maintaining programs that work for the employees to increase rest and decrease fatigue. One concern however, might be whether adequate resources exist to operate trains in this corridor if employees are not able to take a day off at least once every seven days.

Assigned Service
The NS has continued its commitment to developing schedules in which train crews operate in assigned service. Statistics provided from 1997-1998 indicate that 66% of road trains and 95% of yard trains are assigned. Efforts to increase the number of crews operating at assigned times are underway. The Kansas City pilot mentioned above provides employees with a scheduled call time at the home terminal as well as assigned rest days. Meetings have been held in several other locations to increase the number of crews in assigned service on the busiest corridors in the NS system.

One concept being considered on the Cincinnati-Danville run is called preferred pool service. Under this arrangement two crews are assigned to protect two trains, one northbound and one southbound, on an alternating basis of one day on and one day off. Crews operate southbound from their home terminal and northbound from their away-from-home terminal in approximately six hours. Crews in this service lane have a 24-hour off duty period between runs every other day.

Other Scheduling Projects
Several other projects are awaiting ratification at various locations. These involve both an extra board with a one-day rest period and a pool with a seven or eight day rest period with one day off in Birmingham Alabama.

Minimum Undisturbed Rest
As was discussed in previous reports the NS has a policy of minimum undisturbed rest following service. NS has explored this policy since 1995. Most notably, in 1997, the NS instituted a system-wide policy of ten hours undisturbed rest at the home terminal. Some locations have negotiated change in this policy, at away-from-home terminals depending upon the occurrence of short distance runs that might then result in a quicker return to the home terminal. In October 1998, NS implemented a policy of offering employees the option of 24 hours of undisturbed rest after working seven consecutive days. This policy is still in effect. However, it is unclear from material provided whether employees are taking advantage of this policy. In addition, NS has recommended 7:00 A.M. markups for all employees returning to service after a vacation of one week or more. This has become formal policy in several locations.

Lodging Facilities
The Work/Rest committee has also addressed the important topic of lodging facilities. While no new initiatives are underway at this time the committee has addressed this topic and has placed it on its agenda for further review.

Technological Countermeasures
The Work/Rest Committee examined the possibility of utilizing technological countermeasures in the assessment and monitoring of fatigue in locomotive engines. The committee explored the possibility of using the PERCLOS video monitoring technique. However, at one of its meetings decided not to pursue this technology.

NS has made considerable progress in addressing work/rest and fatigue issues on the NS system. As noted above the pilot projects in work/rest scheduling represent a significant increase in activity designed to address fatigue. Moreover, the continuation, with slight modifications of the Elkhart - Chicago IMPAC project, given the extensive research that went in to establishing the program in the first place, is commendable given the operational pressures to perform following the Conrail acquisition. Hopefully, the projects at Buffalo will also be continued. Finally, while considerable progress has been made, discussions about napping while on duty are continuing. Many of the other railroads have adopted a napping policy under certain controlled circumstances. The NS may consider developing a napping policy in the future.

Source: School of Education - University of Denver

The previous two years have seen an increase in the number of projects initiated in an attempt to develop system wide effective fatigue countermeasures in the railroad industry. In order to appreciate the progress that has been made a table with railroads and their performance in the eight areas identified by NARAP has been prepared. As can be seen there have been considerable progress made in the area of napping and education. Most of the Class I railroads have adopted a napping policy. In addition, considerable effort has gone in to improving management and employee knowledge the effects and management of fatigue. There has been somewhat of a retreat from scheduling programs.

Table 12. Countermeasures activities of major railroads.

Counter Measures BNSF UP NS CSX KCSR Amtk CN CP

Scheduling (TY&E)

Minimum Hrs of Rest 10 10 11.5 10 10 10 10 10
Max Hrs of Rest 10, 12, 14 Many Many            
18 hr rest request   Pilot Pilot          
Sites with Time Windows None Pilot None None None None 3 1
Sites with Assigned Days Off 72 100 3 90%        
Percent Scheduled     66% 62%     77%  


Sites Using Naps All All 0 All All   All 1
Napping Facility Sites 0 0 0 0 0   11 0
Technological Countermeasures

(most not considered feasible)

Cab Modifications

(headsets and enhanced seating)

Yes Yes Yes Yes   Yes 25% Yes
Educational Programs

(percent of T&E receiving)

Brochures/Workbooks 100% 100% 100%   100%      
Video Tapes 100% 100% 100% 100%        
Training Seminars 100%   95% 12% 100%      

Sleep Disorder Assessments

Sleep Apnea Screening No Yes   Yes No      

Crew Rest Facilities

Special Facilities       1     16  
Criteria for Lodging Facilities     Yes Yes No     Yes
Lodging Evaluation Program Yes Yes Yes   No   No No

It should be noted that these programs are extremely difficult and complex to implement. Accordingly, most railroads have major pilot projects under study to develop operationally effective methods of managing these fatigue countermeasures. Finally, agreements with assigned days off have been put in place in many locations. However, complex contract negotiations are necessary to develop agreements that satisfy all parties and permit effective implementation.

Labor and Management Issues Regarding Fatigue
The fatigue issue in the railroad is more than just a safety issue. While many would suggest that the science behind the fatigue issues is the driving force it may surprise many to know that critical financial and quality of life issues must be considered when attempting to understand and identify effective fatigue countermeasures.

Fatigue on the railroad is related to financial issues for several reasons. First, over time the BLE and UTU have negotiated contracts that they believe are in the best interest of their membership. In order to maximize earnings the union members have agreed to a contract that places them on 24-hour call to respond to need of the carriers. Over the years, these contracts have been made with the attempt to maximize the earnings potential of the membership. Earnings however, are still for the most part based on a mileage rate of a little over $1 per mile.

Employees who regularly work in designated territories are referred to as "pool" crews. Such crews are generally subject to call at anytime and receive their assignments under the principle of "first in, first out." Local labor and management representatives periodically adjust the number of employees in the pool to prevailing train traffic so that each employee will operate a certain number of miles each month. So, in locations where there is a long distance between terminals the employees are able to receive a high degree of compensation, often with fewer "starts" per month One good example is the run between Spokane, WA and Whitefish, MT, a distance of about 300 miles with minimal congestion. Engineers on this route are able to make the run in under 10 hours, make about $300 per trip, and have a 16 to 20 hour layover. Engineers in Bakersfield, CA, on the other hand, run to Barstow, CA. The route is over rugged terrain and a trip of about 115 miles can take about 9 hours depending on congestion etc. and most also have a 10 to 16 hour layover. So, people in this district must make twice as many trips as those on the longer run in MT to realize the same earnings.

When there is discussion of changing the hours of service rules or altering the schedule to manage fatigue most railroad employees immediately get concerned about the impact on their pocket book.

Life style is another component of the fatigue issue. In many locations the work force is bimodal distributed with respect to age. There are a number of people who have over twenty years of seniority and there are a smaller number with fewer than 5 years of seniority. The demographics of the workforce then consists of one group of individuals who have raised families and gotten their children into college and another group that is younger and just getting started raising families and buying homes. The more senior groups are looking for life style changes that allow for more leisure time while those in the younger groups are looking for earnings potential and more predictable schedules. To the extent that fatigue countermeasures do not interfere with earnings potential, and at the same time promote leisure and family time, they will be supported by the local work force.

At the same time, the nature of the work and the technological changes within the industry have created a change in the ability of the workforce to increase their earnings. The earnings potential of members the BLE-UTU unions is limited to the number of trips that can be made. Over the past 15 years several changes have taken place that have resulted in decreased crew size and longer runs in many locations. On the one hand, technological advances have increased the quality of the locomotive cab environment in the last 10 years. There are a significant number of air conditioned and quieter locomotives. However, the improvements in locomotive efficiency and power, decreased need for large numbers of car movements, the advent of unit trains, etc. have contributed to the pressure to decrease the number of employees and to an increase in the number of trips that can be made.

The situation has gotten to the point where the only way to increase earnings is to increase the number of trips that are being made. Consequently, there are a number of people at various locales that expect to work a considerable amount in order to maintain their lifestyle. These individuals are not very likely to agree to any changes in the work rules or schedule that would essentially decrease the earnings that they have recently gained. Moreover, in some cases, a change in the hours of service could increase the time between trips and therefore decrease the earnings of a particular individual.

Management has also benefited greatly from the technological advances over the past 15 years in terms of improved efficiencies etc. The bottom line has improved due to decreases in crew size and better utilization of equipment. Consequently, an increase in the hours between trips, or a change in the hours of service rules, could result in the need to hire additional employees to meet service demands. So, there are incentives for both management and labor to resist the implementation of fatigue countermeasures that might negatively affect earnings. Regardless of the hours of service act, any fatigue countermeasure that does not favorably impact both parties will face considerable opposition.

The opposition to change is considerable in some locations. Anecdotal reports describe the efforts that union leaders have made trying to move their lodge brothers to a more relaxed attitude and approach to fatigue. In several locations union members have been so opposed to any change in the way that things are done that members have attempted to suppress the adoption of measures to reduce fatigue.

Never the less, some specific proposals have been put forward for schedules that are thought to be improvements in the way that fatigue is managed. One of these proposals involves a schedule of 10 days on and 5 days off per half. The idea is that employees will be able to know when they are going to have days off. Allowing them to plan their leisure and other personal business activities. This approach has been offered instead of the 8-3 approach because it is thought to be less problematic with respect to various agreements.

There are several advantages to the 10-5 approach. As with all scheduling approaches, employees enjoy some level of predictability in their work as well as the ability to spend more time with their families or personal business. In addition, five consecutive days off provides an extended period for employees to recover from any sleep deficit they may have developed over the course of their work period.

Several, problems may also be encountered with the adoption of such a program. First, there is the issue of extended work. Working 10 days in a row could be a challenge to some people. Research by Rosa et. al. (1988) found that performance errors increased after four days on a 12 hour schedule. Heselegrave & Rhodes found that performance errors increased by 15%-18% over a five-day midnight schedule workweek. Second, time off between the shifts may be reduced resulting in an increase in fatigue in a 10-5 schedule. This issue arises in conjunction with the problem of "schedule compression". This term refers to the need to speed up the work-shift cycle in order to be able to get in a sufficient number of trips in the designated time period. Most union locals like to get about 20 trips per month. In order to get in 20 trips in one month, and still only work 20 days, as would be needed in a 10-5 arrangement, it would be necessary to work every day or a least once every 24 hours.

Another issue that local groups face when considering the adoption of the 10-5 schedule from the management point of view has to do with the constraints of contracts. The UTU, BLE and National Carriersí Conference Committee signed an agreement in principal that would make any adoption of solutions designed to address fatigue as cost neutral. This principle reduces the tension of trying new approaches and stems from early pilot projects which provided guaranteed income to employees. The approach, however, may also create difficulty by raising expectations for additional compensation.

Some interpretations of the cost neutral agreement refer to what a person obtains in a pay period, not what the person is able to make over a month or a quarter. With this interpretation concern may arise over some scheduling approaches. For example, depending on the timing of a pay period, an 8-3 schedule might result in an employee working only 8 days in a two-week pay period. While the following pay period would have more paid days, the resulting variability in paychecks may be a problem for some employees. Furthermore, this variability in pay may not meet the criteria of "cost neutral" from the point of view of the union membership or for management either as they are not paying the same amount as is expected. These issues are extremely situation specific to individuals at the local level. Consequently, generalizations are difficult to make.

From the labor perspective, another concern of scheduled work is being on an assignment that has not finished prior to oneís scheduled rest days or receiving a call to work or being called to work prior to a rest day. For example, if one were to get called at 11:45 P.M. when oneís rest day begins at 12 Midnight the length of the trip and layover could eat up the time allotted to rest. Current agreements do not address this issue.

Another important consideration with respect to implementing fatigue countermeasures has to do with the railroad culture of "working when you want". Many people joined the industry with the expectation that they would be able to work as much as they wanted or were physically able to do. This may be an issue for people who are younger. The belief that a union member can pick and choose when and how much they want to work, offers a psychological measure of control that increases satisfaction with railroad work and serves as an important buffer against the negatives associated with the work.

These are some of the most common issues that are raised when considering the implementation of fatigue countermeasures. Both parties believe that it is important to address ways of managing fatigue. However, considerable effort is needed to creatively address the obstacles to implementation.

Best Practices in Fatigue Countermeasures
Having described the obstacles it is imperative that a quick review of the successes be made before closing. Looking across the railroad industry there are a number of key developments that deserve attention and that can be termed best practices. These practices are evident in different railroads and in different locations. They should be considered when planning or developing new and effective fatigue countermeasures.

Assigned Days Off
The practice of giving employees assigned days off after an extended period of work is a very effective method of managing fatigue. This method is perhaps the most effective if a person is going to have extended or variable work schedule. It permits a person to recover from sleep loss and also to plan for family, social and personal needs. Assigned days off should be sufficient in number to allow both recovery from sleep loss, rest, and family, social or personal needs.

The practice of allowing opportunity naps while on duty under predetermined and controlled circumstances is a very effective method of managing fatigue on a short term basis. Napping should not be a substitute for proper rest either at home or on a layover. However, many circumstances can contribute to difficulty in getting rest when it is scheduled, a sick child, an unexpected crisis, other family duties and responsibilities, and circadian rhythms.

Most of the Class I railroads have adopted a policy of napping on the job. This policy gives an employee permission to sleep if necessary. Usually, if a person is very tired it is more beneficial to take a nap than to try to function. It is during these times of intense fatigue that a short nap will enable a person to regain their ability to function effectively and safely.

Currently, various railroads have plans in place to provide anywhere from 1 to 5 assigned days off. The issue of course is balancing a compression of the work cycle to achieve a sufficient number of days off.

Educational Interventions
Clearly the most helpful intervention is for employees to be well educated as to the effects of fatigue and how to manage them. All of the railroads have implemented educational programs of varying types. Seminars, videos, and plenty of readily available brochures are the most common and the most effective means of conveying information to railroad employees.

In addition, more and more anecdotal evidence suggests that in order for specific interventions to be helpful and accepted educational efforts need to be directed towards the individuals at a specific location. These educational activities must assist the local employees in understanding how proposed interventions will work and how they need to be managed off duty.

Work/Rest Consultation
In addition to TY & E employees many non-operating crafts also encounter fatigue challenges. Maintenance of way, engineering, and signal personnel also may find themselves dealing with odd or long hours away from home and possibly in different time zones.

A very effective use of the information obtained from the science of fatigue has been shown to be useful in improving management of work teams. Some evidence from large railroad work projects has shown that describing the effects of shift work and early starts on alertness and performance can be effectively adopted by local management in effectively planning and utilizing work force personnel.

These are just some of the best practices that have been observed in the railroad industry. At this juncture these seem to be the most useful and most easily adopted. Both Cclass I and short-line railroads can easily implement these best practices with little or no difficulty.


News | Contact us | Agreements | Resources | Safety | Site Map | Report A Broken Link

Division 295 - Teamsters Canada Rail Conference - 2016