Parking crashes and Parking Assistance System Design: Evidence from Crash DataBases, the Literature, and Insurance Agent Interviews Paul Green
As shown in Table 15, men were about 15 percent more likely to be involved in parking-related crashes. However, their exposure (were there more male drivers, did they drive and park more often?) was u
Download 0.49 Mb.
|
- Navigate this page:
- Interviews of State Farm Insurance Agents
- Acknowledgments
As shown in Table 15, men were about 15 percent more likely to be involved in parking-related crashes. However, their exposure (were there more male drivers, did they drive and park more often?) was unspecified.Table 15. Number of crashes by year, age, and sex W=women, M=Men, U=Unknown
Impairment did not contribute substantially to parking crashes. Over the 3-year sample, factors that were examined as causing or contributing to crashes included alcohol (2.0 percent), drugs (0.15 percent), distractions (0.25 percent), and cell phone use (0.15 percent). The phone data probably underestimate the current situation as phone use while driving has risen markedly since 2002.As shown in Table 16, most parking crashes occur in daylight hours, though exposure data to determine the relative risk of such crashes were not provided. Table 16. Number of crashes by time of day and year
Figure 2 shows parking-related crashes were evenly distributed among the months of the year, with the fewest crashes (7.2 percent) in February and the most crashes (10.1 percent) in December. The December peak could be due to increased parking associated with holiday-season shopping. 
Figure 2. Percentage of total parking-related crashes (years 2000-2002) In 79.1 percent of parking-related crashes, the first harmful event to occur was also the most harmful event of the crash. Table 17 shows the events associated with the highest crash incidences. Colliding with a motor vehicle in transport (vehicle in motion) accounted for 70 percent of all parking-related crashes between 2000 and 2002, followed by colliding with a parked vehicle (13 percent), and loss of vehicle control (3 percent). Table 17. First harmful event of parking-related crashes
Table 18 shows the consequences of parking-related crashes. Nearly 85 percent of parking-related crashes involved only property damage to the vehicle. The remaining crashes primarily involved injuries to the driver and vehicle passengers, with less than 1 percent of all parking-related crashes involving a pedestrian. Table 18. Parking-related crash implications
Table 19 shows that passenger cars or station wagons accounted for nearly 70 percent of parking-related crashes, followed by pickup trucks with nearly 12 percent, motor homes or vans with about 9 percent, and motorcycles with less than 1 percent. The percentages are fairly stable over the 3-year period. Since the prevalence of each vehicle type was not considered, the data do not necessarily indicate that passenger cars were involved in more accidents than any other type of vehicle. Table 19. Prevalence of vehicle types involved in parking-related crashes
Figure 3 shows where vehicles were struck or struck other vehicles when parking. When entering parking, most damage occurred to the front or passenger side of the vehicle. When leaving parking, nearly one-third of all vehicles had the greatest amount of damage to their left front corner. This could occur when a vehicle is pulling forward out of a parallel parking space and is struck by another vehicle traveling in the roadway. Keep in mind that these data are from primarily road-related crashes, not parking lot crashes, so the number of backing crashes (and rear-end damage) in the database is limited. Also, because these crashes can involve high speeds, there are some rollovers and other crashes for which roof damage will occur. 
Figure 3. Location of greatest vehicle damage from parking-related crashes Table 20 shows the varying extent of damage associated with entering and leaving parking. Leaving parking was associated with fewer instances of slight to no damage than entering parking. However, leaving parked was also associated with more instances of moderate and greater damage, as exiting crashes tended to be more severe. Table 20. Extent of vehicle damage from crashes while entering and leaving parking
Thus, the crash data suggest that parking crashes more commonly involve leaving a space than entering one (by a factor of 2.6), occur slightly more often for men, occur primarily during the daylight hours, rarely involve loss of control, and generally (70 percent) involve no or only slight damage. When there is damage, it is generally property damage (85 percent), with no area on the car more likely to be damaged than another. However, keep in mind that many parking-related crashes are unreported, and these data concern absolute numbers, not risk, where measures of exposure are needed. In terms of engineering solutions, these data suggest the need to focus on supporting leaving from parking spaces, but not to be concerned about lighting (since most crashes occur in the daytime). Interviews of State Farm Insurance Agents Data Collection Method Because many parking crashes occur on private property and are not recorded in public databases, information was sought from those familiar with such crashes, namely insurance agents. In the most states in the U.S., all licensed vehicles must be insured, and even minor crashes often lead to claims.For this project 6 State Farm insurance agents were interviewed about their experience processing claims of crashes related to parking. State Farm is one of the largest automotive insurers in the U.S. and has a strong presence in Michigan. Five of the agents were interviewed in person at their respective offices in the vicinity of Ann Arbor, MI, and a sixth agent from Holland, MI was interviewed by telephone. The Ann Arbor interviews were audiotaped or videotaped. The first agent (actually, the office manager) interviewed served as a pilot subject, and her responses were not included in the final analysis. All subjects were volunteers and not paid for their time. Interviews could last up to an hour each. After having agents complete a consent form, the interviewers inquired about the client and vehicle demographics insured by the agent, and then prompted the agent to provide a narrative of the most recent 2 or 3 parking or low-speed maneuver-related crash claims they could recall. Following the narrative, each agent was asked a predefined set of questions. They included questions about the types of customers they insure (students vs. local residents, ages, professions, U.S. vs. foreign born, etc.) and descriptions of 2 or 3 recent claims that involved parking. Those claims were explored in great detail. Those details concerned a general description of the scenario (where and when the crash occurred, the amount of injury or property damage, the road surface condition, the weather, the visibility and lighting, and the maneuver involved (e.g., turning right into a stall at a particular parking lot). Also obtained were details on the other object. (Did the driver see it? Was it moving? If there was another driver or a pedestrian, was there eye contact? How old the other person was if another person was involved? Was the other person incapacitated by drugs, alcohol, or limited for some other reason? Were there other occupants? What type of vehicle was driven? How could the crash have been presented?) Although the questions were listed in a particular order, the interviews were free flowing and often not answered in the order given. It is important to emphasize that even though a great deal of information was obtained, and the interviewers were careful that the information provided did not uniquely identify an individual. The insured customers about who were described by the agents were not a representative sample of U.S. drivers. Ann Arbor, Michigan, is a college town attended by a large number of foreign students. Although the insured sample has a wide variety of ages, ethnicities, and professions, 3 of the 5 agents indicated that they processed more claims from younger drivers than from other age groups. Additionally, 2 agents stated having more claims from international clients than others. Customers for 3 of the agents primarily drove sedans and sport-utility vehicles, while the other 2 agents insured a reasonable amount of pickup trucks. The majority of agents insured equal numbers of small and large vehicles. All agents indicated that their clients possessed relatively late-model vehicles, with an estimated mean age of less than 5 years. What Were Typical Crashes? Four of the 5 agents described 2 or 3 specific parking crashes, while the remaining agent offered a more general account of those types of crashes. Those 4 agents described a total of 10 crashes involving 20 vehicles, although not every single detail of each crash could be recalled. Some of those crashes were only observed by the agent, and not the subject of an insurance claim they processed. No reported crashes involved pedestrians or injuries. The amount of property damage was highly dependent on the type of vehicles involved and their speed at impact.Although many potential causes were cited, most common were poor parking lot design, human error from inattentiveness, and lack of driver experience. The majority of crashes reported by the agents occurred in parking lots, and 4 of the 10 crashes were directly related to a parking maneuver. Of the 10 crashes, none was associated with entering a parking space, 4 were associated with leaving a parking space, and 6 were associated with low speed maneuvers. Three of 4 of the parking-related crashes involved straight-in parking, while the other crash was a vehicle backing out of angled parking. That trend more likely indicates the greater availability of straight-in parking spaces rather than a greater crash risk. It is interesting that one agent stated, “Angled parking spaces are easier to get in and out of than straight-in parking spaces.” Of the vehicles involved in parking crashes reported by agents, 3 involved contact with a stationary vehicle and 1 with a moving vehicle. The 3 stationary vehicles that were struck were located adjacent to the vehicle leaving a parking space, across the parking lot aisle, and stopped in the parking lot aisle. The remaining crash involved a vehicle moving down the parking lot aisle. All the vehicles involved in low-speed crashes reported by agents were moving forward (Table 21). As noted before, almost all parking-related crashes were caused by a backwards-moving vehicle, and all low-speed crashes involved forward-moving vehicles. Three of the 5 agents personally felt that “backing out” of parking spaces and driveways were the most common scenarios in which a crash would occur. One agent commented: “More people are on ‘auto pilot’ when backing out of driveways.” Table 21. Movement of vehicles involved in crashes
In terms of the driving context, 6 of the 10 crashes occurred on dry pavement, while ice, snow, and gravel was each represented once (Table 22). All combined, the agents recalled the weather conditions of 8 crashes. The majority (7) occurred in clear weather and just 1 low-speed crash occurred in, and was said to be caused by, snow. The light levels were noted as bright in 6 of 8 crashes, as dim in 1 crash, and as night in another. Table 22. Road surface condition during crash occurrence
The agents reported that the majority of crashes involved 2 vehicles, rather than a vehicle and a fixed object. Who Were the Drivers?
Of the 19 drivers, 16 were recalled, 8 men and 8 women. Of them, the ages were recalled for 13 drivers: 8 were 16 to 28, 4 were 38 to 50, and 1 was 70. In the 10 crashes, no driver was incapacitated by medication, drugs, alcohol, or physical disability.In 4 of the 10 instances, agents felt people drove too fast given their experience and the road conditions. Agents recalled the sight conditions for only 8 of the crashes, and in 5 of them the driver could see the other vehicle. There was only 1 reported case of eye contact between drivers. HOW COULD PARKING CRASHES BE PREVENTED?Most agents believed the main problems were the design of parking facilities and human errors deriving from inexperience, inattentiveness, and non-compliance with the basic principles taught in driver’s training. When prompted, the agents offered the following ideas to help drivers avoid parking-related crashes: A device to help the driver see behind their vehicle without having to back out too far Large round mirrors installed in parking lots and driveways to provide drivers a better view of their surroundings A device that beeps to indicate the presence of vehicles within a 100-foot radius around the vehicle (it would activate only while the car is moving and not in heavy traffic on a roadway) A camera that provides a panoramic view of the rear of the vehicle Sensors to detect and alert drivers to vehicles in the driver’s blind spots CONCLUSIONS The conclusions that follow are based on all 3 sets of data, none of which is perfect, and most of which focus on parking crashes, the primary crash type of interest. Many of the studies are 20 or more years old, representing times when cars predominated the vehicle fleet and traffic density was much lower. The current U.S. mix of cars, minivans, SUVs, and light trucks can make seeing other vehicles (and the boundaries of one’s own vehicle) quite difficult in parking maneuvers. Accordingly, the problems and consequent frequency of crash types may have also changed. Secondly, the Michigan crash data, a representative sample of police-reported crashes for a state with geography and a vehicle fleet typical of the U.S., primarily concerns crashes on public roads, not private property such as parking lots and garages. Hence, some of the desired crash data is missing from that set. Finally, while the information from insurance agents was useful, the sample size was small. However, in spite of these limitations, the findings from each of these sources are mutually reinforcing. 1. Depending on the source and data set, anywhere from 1/2 to 3/4 of all parking crashes involve backing from a parking spot, primarily from a parking stall. Most commonly, the crash involves striking another vehicle moving in the aisle, not a parked car. 2. Crashes at parking lot aisle intersections generally occur because drivers are moving too quickly, but sometimes because there are no stop signs to regulate traffic flow. 3. When crashes involve fixed objects, a building is the most commonly struck object. 4. The literature suggests that crash rates are higher for 8-1/2-foot stalls than for wider stalls, but the data are from a time when vehicles were much smaller and more likely to be cars. 5. Most parallel parking crashes occur on major streets. 6. Men account for a slightly greater fraction of parking crashes than women, but it is uncertain if there is a difference when considering exposure (e.g., miles driven, number of parking events). 7. No age group is more likely to have parking crashes when the frequency of parking crashes is compared with their overall crash rate. Keep in mind that very young drivers are relatively more likely to be involved in all types of crashes including parking-related crashes. 8. Impairment due to alcohol and drugs is a relatively minor factor in parking crashes. 9. Lighting does not seem to be a factor in parking crashes. Day/night differences seem to reflect exposure to those conditions. 10. There are few month-to-month differences in parking crashes, except that more crashes occur in December than any other month, possibly due to shopping-related parking during the holiday season. 11. When backing up, the most common place to look is over the right shoulder. The literature, the crash data from Michigan, and insights from the insurance agents all point to the potential value of a system that would allow drivers to see down a parking aisle when backing out of a stall into an aisle or street. The system should provide visual or other types of information to the rear and sides of the vehicle for backing from both perpendicular and angled stalls. Often the driver cannot see around other vehicles, in particular vans, minivans, light trucks, and SUVs, which, as a group, comprise over 50 percent of the new vehicles sold in the U.S. A secondary opportunity is assisting with parallel parking on major (presumably busy) streets, though there is less evidence pertaining to that situation. In addition, the data suggest that studies should certainly include young drivers (who are most likely to be involved in crashes) and evaluate 8-1/2-foot-wide stalls for perpendicular and angle parking (since there are more crashes for that width), but do not need to consider impairment due to drugs or alcohol. Though these findings should hold for most of North America, it is uncertain how differences in traffic density, the types of parking available, the speeds on public roads, the sizes of vehicles available, geography, and even driver aggressiveness, are likely to influence the application of these results to South and Central America, Europe, or Asia. REFERENCES Box, P. C. (1981). Parking Lot Accident Characteristics. ITE Journal, 51(12), December, 12-15. Box, P. C. (2001). Angle parking issues revisited. ITE Journal, 72(3), March, pp. 36, 43-47. Cullinane, B. Smith, D. and Green, P. (2005). Where, When, and How Well People Park: A Phone Survey and Field Measurements (Technical Report UMTRI-2004-18), Ann Arbor, MI: University of Michigan Transportation Research Institute. Garwood, F. (1959). Fatal and Serious Accidents Involving Parked Vehicles (1958) (Research Note No RN/3568/FG 1958), Crowthorne, England: Road Research Laboratory. Gadgil, S. and Green, P. (2005). How Much Clearance Drivers Want While Parking: Data to Guide the Design of Parking Assistance Systems, Proceedings of the Human Factors and Ergonomics Society 49th Annual Meeting, Santa Monica, CA: Human Factors and Ergonomics Society, 1935 –1939 (CD-ROM). Green, P., Gadgil, S., Walls, S., Amann, J., and Cullinane, B. (2004). Desired Clearance Around a Vehicle while Parking and for Low Speed Maneuvers (Technical Report 2004-30). Ann Arbor, MI: University of Michigan Transportation Research Institute. Green, P., Gillespie, T., Reifeis, S., Wei-Haas, L., and Ottens, D. (1984). Subjective Evaluation of Steering Effort Levels (Technical Report UMTRI 84-39), Ann Arbor, MI: University of Michigan Transportation Research Institute (NTIS PB 86 135019/AS). Harpster, J. Huey, R. and Lerner, N. (1996). Field Measurement of Naturalistic Backing Behavior, Proceedings of the 39th Annual Meeting of the Human Factors and Ergonomics Society, Santa Monica, CA: Human Factors and Ergonomics Society, pp. 891-895. Harpster, J. L., Huey, R. W., Lerner, N. D., and Steinberg, G. V. (1996). Backup Warning Signals: Driver Perception and Response (Technical Report DOT/HS 808 536), Washington, D.C.: U.S. Department of Transportation, National Highway Traffic Safety Administration. Humphreys, J. B., Box, P. C., Sullivan, T. D., and Wheeler, D. J. (1978). Safety Aspects of Curb Parking (Technical Report FHWA-RD-79-76), Washington, D.C.: U.S. Department of Transportation, Federal Highway Administration. McCoy, P. T., Ramanujam, M., Moussavi, M., and Ballard, J. L. (1990). Safety Comparison of Types of Parking on Urban Streets in Nebraska, Transportation Research Record, No. 1270, pp. 28-41. Rubin, R. and Green, P. (2005). Design Guidelines for Video-Based Parking Assistance Systems (Technical Report UMTRI 2005-9), Ann Arbor, MI: University of Michigan Transportation Research Institute. Seburn, T. J. (1965, Estimated date). Relationship between Curb Uses and Traffic Accidents, New Haven, CN: Yale University. Smith, D., Green, P., and Jacob, R. (2004). Parking and Low-Speed Crashes: Crash Database, Literature, and Insurance Agent Perspectives (Technical Report UMTRI 2004-9), Ann Arbor, MI: University of Michigan Transportation Research Institute. Walls, S.M., Amann, J., Cullinane, B., Green, P., Gadgil, S., and Rubin, R. (2004). Alternative Images for Perpendicular Parking: A Usability Test of a Multi-Camera Assistance System (Technical Report UMTRI 2004-17), Ann Arbor, MI: University of Michigan Transportation Research Institute. Walls, S.M., Green, P., Gadgil, S., Amann, J., and Cullinane, B. (2004). Alternative Images for Parallel Parking: A Usability Test of a Multi-Camera Assistance System (Technical Report UMTRI 2004-31), Ann Arbor, MI: University of Michigan Transportation Research Institute. Acknowledgments This research was supported by Nissan Motor Company whose support is gratefully acknowledged. Dan Smith (now with Toyota) and Renju Jacob provided significant contributions to the original technical report on which this paper is based. CONTACT Paul Green University of Michigan Transportation Research Institute (UMTRI) Human Factors Division 2901 Baxter Road Ann Arbor, Michigan 48109-2150 USA pagreen@umich.edu www.umich.edu/~driving Directory: ~driving -> publications publications -> Methods for assessing the safety and usability of in-vehicle systems: Lessons from umtri driver interface research Download 0.49 Mb. Share with your friends:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||