Creaser, J.I., Caird, J.K., Edwards, C., & Dewar, R.E. (2002). A qualitative analysis of driver distraction and intentional actions at railway grade crossing accidents in Canada. Joint Conference of the Association of Canadian Ergonomists and Institute of Industrial Engineers [CD-ROM]. Association of Canadian Ergonomists. A QUALITATIVE ANALYSIS OF DRIVER DISTRACTION AND INTENTIONAL ACTIONS IN RAILWAY GRADE CROSSING ACCIDENTS IN CANADA CREASER JANET, CAIRD JEFF, EDWARDS CHRISTOPHER Cognitive Ergonomics Research Laboratory, Department of Psychology, University of Calgary, 2500 University Dr. N.W., Calgary, AB T2N 1N4, [email protected] DEWAR ROBERT Western Ergonomics, Calgary, Alberta A qualitative analysis using key word search techniques of the Transportation Safety Board of Canada’s (TSB) Rail Occurrence Database System (RODS) identified incidents where distraction or intentional actions contributed to railway grade crossing accidents. Based on the literature review, search parameters were developed and tested on the 7,776 accident narratives logged in RODS between January 1, 1983 and November 7, 2001. Only narratives logged between January 1, 1990 and November 7, 2001 were sufficiently detailed to be subjected to a key word search. There were 86 narratives that indicated an intentional driver action and 39 narratives that indicated driver distraction as crash contributors. An additional 31 narratives indicated distraction or an intentional action as a potential factor. Behaviours cited were similar to those found in studies of railway grade crossing accidents in other countries, and multiple factors were often present. Overall conclusions are limited due to the small number of narratives found for each category. Key words: Railway-grade crossing accidents, driver behaviour, human error UNE ANALYSE QUALITATIVE DE CONDUITE DISTRAITE ET ACTIONS INTENSIONNES DANS LES ACCIDENTS AUX PASSAGES A NIVEAU Une analyse utilisant une technique de recherche avec des mots clés du Bureau de la securité des transports du Canada (BST) Rail Occurrence Database System (RODS) oú des incidents de distraction, ou des actions intetionnelles aurait contribuées a des accidents aux passages a niveau. Fondé sur la littérature, la recherche a été developée et examinée sur les 7,776 recits d’accidents enregistrees a RODS entre le premier janvier 1983 et le 7 novembre 2001. L’examen initial des récits a indiqué que seulement des récits notés entre le premier janvier 1990 et le 7 novembre 2001 ont été suffisamment détaillés pour être soumis a une recherché de mot clé. Il y avait 86 récits qui indiquaient une action intentionnelle du conducteur et 39 qui indiquaient un conducteur distrait comme causes d’accidents. 31 autres récits indiquaient aussi que la cause de l’accident aurait pu être soit la distraction ou des actions intentionnelles. Les comportement observés étaient semblables à ceux trouvés ressemblaient dans d’autres études d’accidents aux passages à niveau dans d’autres pays. Parce-que les données ne sont pas completes les conclusions seronts limitées. Mots cles : Accidents au passages à niveau, comportement de conducteur

INTRODUCTION From 1991 to 2001, an average of 316 highway-railway grade crossing accidents and 38 fatalities (i.e., vehicle occupants) have occurred each year in Canada. Although many factors contribute both individually and in combination to these accidents, distraction and intentional actions at crossings are two common driver-related factors (1,8,11). Distraction can be internal or external to the driver. Internal distraction is defined as being distracted by internal cognitive processes, such as daydreaming, worrying, or being excited about an event in one’s life (16). External distraction occurs when objects or events outside of the vehicle distract a driver (14), such as oncoming traffic. In-vehicle distractions, such as attending to children or answering a cell phone, also pose a threat to drivers. When distracted, drivers often fail to see railway crossing signs, crossings, or trains, particularly at passive crossings (i.e., protected by advance warning signs only). However, other factors may also prevent the driver from seeing signs, crossings or trains, such as obstructed sight lines or adverse weather conditions (2). Distraction can also occur when a crossing is near to a road-road intersection in that a driver’s attention is focused on navigating the intersection and not on the state of the tracks beyond. Active crossings with bells, lights and gates are better at attracting attention, but produce other forms of driver error. Drivers will often violate crossing signals by driving around gates or through crossings with activated flashing lights (1,2). In the context of a highway-railway grade crossing accident, when a driver ignores the active and passive signals and signs, often in an attempt to beat the train, the act can be interpreted as an intentional violation (12). In one observational study, 68% of drivers observed violating an active crossing said they used the crossing at least four times a week and 19% said they used the crossing two to four times a week. Most respondents said they ignored the signals or went around barriers because the “train was not in sight” or the “train was stopped for an unreasonable amount of time” (1). These responses agree with other findings that show drivers required to wait for long time periods at crossings before trains arrive are more likely to violate active crossing signals (2,8). Finally, although deliberate attempts to beat a train can be considered violations, drivers may intentionally choose to enter crossings without committing a violation. For example, at passive crossings, drivers may misjudge how far away and how fast a train is travelling and continue to cross. Perceptual misjudgements of train speed and distance are well documented (10). Fatigue, impairment due to drugs or alcohol, and driving too fast for the conditions can also be considered intentional actions or violations that lead to accidents. The qualitative assessment described in this paper was conducted as part of a larger study of railway grade crossing accident contributors in Canada (4). A key word search of accident narratives from the Transportation Safety Board of Canada’s (TSB) Rail Occurrence Database System was used to expand the understanding of driver behaviour at railway grade crossings. The influence of driver distraction and intentional actions, such as crossing against signals, were of particular interest. Qualitative analyses of narratives have yielded interesting results when applied to the use of in-vehicle devices such as cell phones (6) and driver distraction (15). This method provides a descriptive look at factors that contribute to accidents, such as specific driver behaviours. To our knowledge, the application of this method to highway-railway grade crossing accidents has not been done previously.

METHODS Narratives were extracted from the 7,776 accidents entered in RODS as of November 7, 2001. A key word search was performed using the broad categories of distraction, intentionality, visibility/weather, fatal, second train, alcohol-related and intersection-related. Of the 7,776 crossing accidents in RODS, 6,402 had a narrative. Of the missing 1,374 narratives, 1,307 were missing from the years 1983-1989. Only 67 narratives were missing for the years 1990-2001. Narratives logged before 1990 were not detailed enough to be used in this study. As a result, accident narratives (N = 3,990) from 1990-2001 were used in the qualitative analysis. Several methods were used to generate the final list of key words, including test text searches, consultation with related literature and thesaurus look up. The frequency of accidents in any one category is not necessarily representative of the distribution of accidents at highway-railway grade crossings and conclusions regarding the rank ordering of types of narratives should be cautiously interpreted. RESULTS Intentional actions There were 86 narratives that indicated a driver acted intentionally in a manner that may have contributed to an accident. In 35 cases, drivers went around activated gates and in four cases drivers went around both stopped cars and gates. In 10 cases at flashing light only crossings, drivers went around stopped or slowing vehicles at the crossing. In 16 cases, the narrative noted that drivers seemed to have attempted to beat the train over the crossing. However, cases where a vehicle slows, then proceeds (10 cases) in front of a train may actually reflect the normal behaviour of drivers at crossings. That is, drivers tend to slow down on grade crossing approaches (9), and drivers assumed to be intentionally attempting to beat the train might have simply failed to see the lights or advance warning signs on the approach, either due to distraction or other factors. Therefore, these 16 cases are difficult to classify. There were only five narratives found that mentioned alcohol as a contributing factor. Since 1998, alcohol can also be queried as an unsafe condition in a RODS report, but there were only 16 accidents with alcohol listed as an unsafe condition between 1998-2001. Three narratives indicated fatigue as a factor. In 25 accidents fog was a factor and in eight snow was a factor, but it was unknown whether drivers were going too fast for the conditions. Distraction-related accidents Thirty-nine narratives identified driver distraction as a contributor to a crossing accident. In 19 of these cases, drivers saw the train when it was too late to stop in time (N = 9) or failed to detect either the train or signals at all (N = 10). Sixteen of these accidents occurred at passive crossings, with the other three occurring at crossings with flashing lights and bells, but no gates. Other factors, such as visibility restrictions, could also explain these accidents, however, the drivers did not provide reasons for why they did not see the crossing in time. In seven narratives, the use of a cellular phone was indicated as a contributing factor. Four narratives indicated the presence of an internal distraction and three indicated the presence of an external distraction. Three narratives indicated the drivers were distracted by a conversation with passengers and one narrative indicated that the driver was distracted while attempting to adjust his radio. Accidents near an intersection and second-train accidents Thirty-one narratives noted that a driver was required to navigate an intersection just before reaching the crossing. Although advance warning signs in Canada exist to indicate the

presence of a crossing near an intersection, whether drivers ignore these signs, do not comprehend them or simply fail to see them is not known. Whether distraction was a factor in these accidents is not definitively known, however, they are of interest as it has been suggested that an intersection before a crossing may distract attention away from the crossing (16). Ten narratives indicated that a second train was involved in the accident. This type of accident occurs when a driver assumes the tracks to be clear after one train passes and is struck by a second oncoming train after entering the crossing. At active crossings, the driver usually violates signals, not realizing they are still activated for a second train rather than the one that just passed. DISCUSSION Overall, the narratives provided some detailed descriptions of how driver distraction and intentional actions may contribute to railway grade crossing accidents. However, the prevalence of distraction and intentional actions in crossing accidents cannot be sufficiently determined from this sample. The small number of returned narratives (156 out of a possible 3,990) possibly reflects reporting practices within the TSB because driver behaviour and highway safety is not within the TSB’s mandate. Most narratives only contain short descriptions and do not necessarily focus on citing a possible cause, such as driver error. Despite the small number of cases for each category, clear examples of distraction and intentional actions were found, such as cell phone use, attending to oncoming traffic or attempting to beat a train. Furthermore, the boldness of certain violations was apparent, with drivers entering the oncoming lane of traffic to go around gates and vehicles. That countermeasures specifically designed to protect drivers, such as gates and lights, are commonly violated might indicate that drivers do not fully comprehend the dynamics of railway crossings and therefore, underestimate the potential for a crash. Neither the fatigue nor alcohol searches yielded good results. Fatigue is thought to be a major contributor to vehicle accidents, but identifying its presence when an accident occurs is problematic (3). The frequency of alcohol-related incidents was expected to be greater because grade crossing accidents are considered to have alcohol involvement rates similar to those for all fatal highway crashes (7). Although searches turned up weather-related factors (e.g., ice, fog, snow), none of the narratives suggested that a driver was going too fast for the conditions—a situation that may contribute to not seeing crossing signs, signals or a train in time to stop. Although narratives were categorized using only one factor, multiple factors related to the driver and the crossing were present in several narratives. For example, one narrative was considered as internal distraction because the driver was late for an event when the accident occurred. However, other salient factors were also present. The following is the narrative as it appeared in RODS, but with modifications to remove identifying details: “The vehicle was travelling eastbound on the south side road parallel to the track. The driver did not stop at the intersection as required by the stop sign, but turned north onto the crossing and was struck on the driver’s door by the train. A witness heard the train whistle and she saw the driver look up the hill to the north and then look to the south before he turned north onto the track. They were late for a parade. The driver lived within two blocks of the crossing. The driver and his girlfriend were not wearing seatbelts, but the two boys were wearing seat belts.” First, the driver could have been distracted by a desire to reach the parade on time and, second, by the interactions between the three passengers. Third, being familiar with a crossing does not necessarily reduce one’s risk of having an accident, especially if trains are infrequent on the crossing (11,16). Finally, the driver was required to stop at a stop sign before he turned across the tracks. From the narrative, it appears the driver may have been looking for traffic to his left and right (north and south), but failed to stop at the stop sign. It is

not clear if he deliberately violated the stop sign, but violations compounded by additional driver errors are a recipe for disaster (12). The relative importance of any one of these four factors in isolation or together is difficult to determine because the driver was killed. Several other narratives showed similar combinations of factors, such as distraction and violations. In conclusion, incidents of driver distraction and intentional actions as contributors to railway grade crossing accidents were evident in the RODS narratives. The narratives extracted provided a brief but descriptive glimpse of how these factors possibly contribute to crossing accidents. However, overall conclusions cannot be drawn due to the data limitations and the classification of accidents in only one category. Preventing violations and ameliorating the effects of distraction are complex issues, but certain countermeasures are available now and others will be available in the future to address these problems. For example, median barriers are a simple, cost effective method to prevent drivers from going around gates (5) and new technologies may be able to warn drivers in advance that they are approaching a grade crossing (13). The costs and benefits of countermeasures are currently being investigated in the U.S. and may offer insight into preventing accidents in Canada. REFERENCES (1) Abraham, J., Datta, T.K., Datta, S. (1998). Driver behavior at rail-highway crossings. Transportation Research Record 1648, 28–34. (2) Berg, W.D., Knoblauch, K., Hucke, W. (1982). Causal factors in railroad-highway grade crossing accidents. Transportation Research Record 847, 47–54. (3) Brown, I. D. (1994). Driver fatigue. Human factors, 36, 219–231. (4) Caird, J.K., Creaser, J.I., Edwards, C.J., Dewar, R.E. (2002). A human factors analysis of highway-railway grade crossing accidents in Canada. Montreal, Canada: Transport Development Centre: Transport Canada. (5) Carroll, A.A., Haines, M. (2002). North Carolina “sealed corridor” phase I safety assessment. Transportation Safety Board [CD-ROM]. Washington, D.C.: TRB. (6) Goodman, M.J., Tijerina, L., Dents, D.R., Wierwille, W.W. (1999). Using cellular telephones while driving: Safe or unsafe. Transportation Human Factors, 1(1), 3-42. (7) Klein, T., Morgan, T., Weiner, A. (1994). Rail-highway crossing safety fatal crash and demographic descriptors (Pub. No. DOT HS 808 196). Washington, D.C.: U.S. Department of Transportation, National Highway Traffic Safety Administration. (8) Meeker, F., Fox, D., Weber, C. (1997). A comparison of driver behavior at railroad grade crossings with two different protection systems. Accident Analysis and Prevention, 29, 11-16. (9) Moon, Y.J., Coleman, F. (1999). Driver’s speed reduction behavior at highway rail th intersections. In the Paper reprints of the 78 Annual Meeting of the Transportation Research Board [CD-ROM]. Washington, D.C.: TRB. (10) Mortimer, R. G. (1988). Human factors in highway-railroad grade crossing accidents. In G.A. Peters & B.J. Peters (Eds.), Automotive engineering and litigation (Vol. 2, pp. 35–69). New York: Garland Law. (11) National Transportation Safety Board (1998). Safety study: Safety at passive grade crossings, Volume 1: Analysis (PB98-917004, NTSB/SS-98/02). Washington, D.C.: NTSB. (12) Reason, J. (1997). Managing the risks of organizational accidents. Brookfield: Ashgate. (13) Richards, H.A., Bartoskewitz, R.T. (1995). The intelligent highway-rail intersection integrating ITS and ATCS for improved grade crossing operation and safety. In Safety of Highway-Railroad Grade Crossings: Research Needs Workshop, Volume II – Appendices. (DOT/FRA/ORD-95/14.2). Washington, D.C.: Department of Transportation, Federal Railroad Administration. (14) Treat, J.R. (1980). A study of the precrash factors involved in traffic accidents. The HSRI Review, 10(1), 1–35.

(15) Wierwille, W.W., Tijerina, L. (1996). An analysis of driving accident narratives as a means of determining problems caused by in-vehicle visual allocation and visual workload. In A.G. Gale et al. (Eds.). Vision in Vehicles – V. (pp. 79-86). Elsevier Sciences B.V. (16) Wigglesworth, E.C. (1979). The epidemiology of road-rail crossing accidents in Victoria, Australia. Journal of Safety Research, 11(4), 162–171.