Cooperative warning systems: The impact of false and unnecessary alarms on drivers’ compliance

Cooperative warning systems have a great potential to prevent traffic accidents. However, because of their predictive nature, they might also go along with an increased frequency of incorrect alarms that could limit their effectiveness. To better understand the consequences associated with incorrect alarms, a driving simulator study with N = 80 drivers was conducted to investigate how situational context and warning urgency jointly influence drivers’ compliance with an unreliable advisory warning system (AWS). The participants encountered several critical urban driving situations and were either assisted by a 100% reliable AWS, a 60% reliable AWS that generated false alarms (without obvious reason) or a 60% reliable AWS that generated unnecessary alarms (with plausible reason). A baseline drive without any assistance was also introduced to the study. The warnings were presented either only visually or visual-auditory. In line with previous research, drivers’ compliance and effectiveness of the AWS was reduced by false alarms but not by unnecessary alarms. However, this so-called cry wolf effect (Breznitz, 1984) was only found in the visual-auditory condition, whereas there was no effect of warning reliability in the condition with visual AWS. Furthermore, false but not unnecessary alarms caused the participants to rate the AWS less favourably during a follow-up interview. In spite of these negative effects of false alarms, a reduction in the frequency of safety-critical events (SCEs) and an earlier braking onset were evident in all assisted drives compared with that of non-assisted driving, even when the AWS was unreliable. The results may thus lower concerns about the negative consequences of warning drivers unnecessarily about upcoming traffic conflicts if the reasons of these alarms are comprehensible. From a perspective of designing AWS, we recommend to use less urgent warnings to prevent the cry wolf effect.     

Drivers’ eye movements as a function of collision avoidance warning conditions in red light running scenarios

The intersection collision avoidance warning systems (ICAWSs) have substantial potentials in improving driving performance and reducing the number and severity of intersection collisions, through helping drivers timely detect hazardous conflicting vehicles in precrash scenarios. However, the influences of ICAWS on drivers’ visual performance have barely been discussed. This study focuses on exploring the patterns in drivers’ eye movements as a function of ICAWS’s warning conditions in red light running scenarios based on a driving simulation experiment. Two types of speech warning conditions including warning timings (varied form 2.5 s to 5.5 s) and directional information (with or without) are examined, and the no-warning condition is the baseline. The results revealed that more subjects would be likely to benefit from the ICWAS under the earlier warning timings. The warning condition of 4.5 s ahead of a collision had the best effectiveness in terms of visual performances. Under such a warning timing, drivers had shorter fixation duration and higher frequency of searching for the red light running (RLR) vehicles. Compared to the warning condition without directional information, the directional warning information could capture drivers’ attention more efficiently, help driver direct fixations toward the RLR vehicles more quickly and lead to more scanning activities. Compared to female drivers, male drivers had more scanning activities when approaching intersections, detected the RLR vehicles more quickly and were more likely to avoid the RLR collisions. Besides, the experiment results indicated that the female drivers were more inclined to trust the warning information and got more benefits from the RLR-ICAWS in terms of the crash risk reduction rate than male drivers. Finally, the conclusions lead the way toward warning condition design recommendations for improving the effectiveness of the RLR-ICAWSs.     

Behavioral adaptation of young and older drivers to an intersection crossing advisory system

An advanced driver assistance system (ADAS) provided information about the right of way regulation and safety to cross an upcoming intersection. Effects were studied in a longer-term study involving 18 healthy older drivers between the ages of 65 and 82 years and 18 healthy young drivers between the ages of 20 and 25 years. Participants repeatedly drove 25 km city routes in eight sessions on separate days over a period of two months in a driving simulator. In each age group, participants were randomly assigned to the control (no ADAS) and treatment (ADAS) group. The control group completed the whole experiment without the ADAS. The treatment group drove two sessions without (sessions 1 and 7) and six times with ADAS. Results indicate effects of ADAS on driving safety for young and older drivers, as intersection time and percentage of stops decreased, speed and critical intersection crossings increased, the number of crashes was lower for treatment groups than for control groups. The implications of results are discussed in terms of behavioral adaptation and safety.     

Jerky driving--An indicator of accident proneness?

This study uses continuously logged driving data from 166 private cars to derive the level of jerks caused by the drivers during everyday driving. The number of critical jerks found in the data is analysed and compared with the self-reported accident involvement of the drivers. The results show that the expected number of accidents for a driver increases with the number of critical jerks caused by the driver. Jerk analyses make it possible to identify safety critical driving behaviour or “accident prone” drivers. They also facilitate the development of safety measures such as active safety systems or advanced driver assistance systems, ADAS, which could be adapted for specific groups of drivers or specific risky driving behaviour.     

Adapting ISA system warnings to enhance user acceptance

Inappropriate speed is a major cause of traffic accidents. Different measures have been considered to control traffic speed, and intelligent speed adaptation (ISA) systems are one of the alternatives. These systems know the speed limits and try to improve compliance with them. This paper deals with an informative ISA system that provides the driver with an advance warning before reaching a road section with singular characteristics that require a lower safe speed than the current speed. In spite of the extensive tests performed using ISA systems, few works show how warnings can be adapted to the driver. This paper describes a method to adapt warning parameters (safe speed on curves, zone of influence of a singular stretch, deceleration process and reaction time) to normal driving behavior. The method is based on a set of tests with and without the ISA system. This adjustment, as well as the analysis of driver acceptance before and after the adaptation and changes in driver behavior (changes in speed and path) resulting from the tested ISA regarding a driver's normal driving style, is shown in this paper. The main conclusion is that acceptance by drivers increased significantly after redefining the warning parameters, but the effect of speed homogenization was not reduced.     

Driving performance assessment: effects of traffic accident location and alarm content

According to accident statistics for Taiwan, the two most common traffic accident locations in urban areas are roadway segments and intersections. On roadway segments, most collisions are due to drivers not noticing the status of leading vehicle. At intersections, most collisions are due to the other driver failing to obey traffic signs. Using a driving simulator equipped with a collision warning system, this study investigated driving performance at different accident locations and between different alarm contents, and identified the relationship between crash occurrences and driving performance. Thirty participants, aged 20-29 years, were recruited in this study. Driving performance measures were perception-reaction time, movement-reaction time, speed and a crash. Experimental results indicated that due to different demands for processing information under different traffic conditions, driving performance differed at the two traffic accident locations. On a roadway segment, perception-reaction time for a beep was shorter than the time for a speech message. Nevertheless, at an intersection, a speech message was a great help to drivers and, thus, perception-reaction time was effectively reduced. In addition, logistic regression analysis indicates that perception-movement time had the greatest influence on crash occurrence.     

Tactile warning signals for in-vehicle systems

The last few years have seen growing interest in the design of tactile warning signals to direct driver attention to potentially dangerous road situations (e.g. an impending crash) so that they can initiate an avoidance maneuver in a timely manner. In this review, we highlight the potential uses of such warning signals for future collision warning systems and compare them with more traditional visual and auditory warnings. Basic tactile warning signals are capable of promoting driver alertness, which has been demonstrated to be beneficial for forward collision avoidance (when compared to a no warning baseline condition). However, beyond their basic alerting function, directional tactile warning signals are now increasingly being utilized to shift the attention of the driver toward locations of interest, and thus to further facilitate their speeded responses to potential collision events. Currently, many researchers are focusing their efforts on the development of meaningful (iconic) tactile warning signals. For instance, dynamic tactile warnings (varying in their intensity and/or location) can potentially be used to convey meaningful information to drivers. Finally, we highlight the future research that will be needed in order to explore how to present multiple directional warnings using dynamic tactile cues, thus forming an integrated collision avoidance system for future in-vehicle use.     

Effectiveness evaluation of simulative workshops for newly licensed drivers

The current study set to examine the effects of simulator use in driving instruction on newly licensed drivers, comparing the road safety knowledge and reported intended behavior, as well as the actual driving performance of new drivers. Participants consisted of 280 newly licensed driver, of which 140 whose drivers license training included additional simulator-based lessons, and 140 drivers whose training precluded simulator-based lessons. All drivers answered questionnaires pertaining to their intended safe driving behaviors (according to Ajzen's (2000) theory of planned behavior), and to their traffic safety knowledge. Of the initial sample, 40 drivers received actual driving performance evaluation by an expert driving instructor, as well as by in-vehicle data recorders (IVDRs). We assumed that safer drivers report safer driving intentions, demonstrate greater traffic safety knowledge, evaluated as safer drivers by the driving instructor, and display lower and stable driving parameters on the IVDRs. We hypothesized that theoretical driving studies combined with practical training on simulators will elevate the safety level of novices driving. Hierarchical regression analyses on driving intentions indicated that drivers who did not receive simulator-based lessons demonstrated safer driving intentions compared to drivers who received simulator-based lessons. This pattern possibly indicating the drivers who received simulator-based lessons felt more confident in their driving abilities compared to drivers who did not receive simulated training. No significant difference was found in traffic safety knowledge, or in the evaluation of the expert driving instructor. IDVR data comparisons indicated drivers who received simulator-based lessons braked more often and were less prone to headway events, suggesting a more responsive driving style. These findings do not point to any significant advantage or disadvantage of the current simulator-based driving training over other driving training methods.     

Driving while conversing: Cell phones that distract and passengers who react

The research systematically compared the driving performance and conversational patterns of drivers speaking with in-car passengers, hands-free cell phones, and remote passengers who could see the driver's current driving situation (via a window into a driving simulator). Driving performance suffered during cell phone and remote passenger conversations as compared with in-car passenger conversations and no-conversation controls in terms of their approach speeds, reaction times, and avoidance of road and traffic hazards. Of particular interest was the phenomenon of conversation suppression, the tendency for passengers to slow their rates of conversation as the driver approached a hazard. On some occasions these passengers also offered alerting comments, warning the driver of an approaching hazard. Neither conversation suppression nor alerting comments were present during cell phone conversations. Remote passengers displayed low levels of alerting comments and conversation suppression, but not enough to avoid negative effects on driving performance. The data suggested that conversation modulation was a key factor in maintaining driving performance and that seeing the road and traffic was not sufficient to produce it. A second experiment investigated whether a cell phone modified to emit warning tones could alleviate some of the adverse effects typically associated with cell phone conversations. The modified cell phone produced discourse patterns that were similar to passenger conversations and driving performance nearly as good as that of drivers who were not conversing. This latter finding supported the argument that conversation modulation is a key ingredient in avoiding adverse effects of conversations with drivers, rather than the physical presence of an in-car passenger.     

自动驾驶接管中的分层振动警示设计研究

目的驾驶员在自动驾驶系统遇到无法处理的交通情况时需要尽快恢复情景意识,进行驾驶操控权的接管,由于自动驾驶中驾驶员将更多的从事驾驶次任务,需要进行有效的自动驾驶接管信息提示设计,降低认知负荷,提高接管绩效,改善驾驶体验。方法触觉显示是在视听通道失效的情况下,获取驾驶情景信息的有效方式。基于自主开发的触觉坐垫,研究分层振动警示接管策略对情景意识恢复和用户体验的影响。实验共有24名被试,每个被试将进行单层级、双层级和三层级接管实验。被试在接收到振动警示后,需要中止当前的非驾驶相关任务,接管汽车的控制,并躲过风险情境。结论实验结果表明,相对于单层级振动警示,多层振动警示能提高接管绩效,尤其在高负荷道路和低负荷道路中。三层级振动警示的综合评价最高,但同时也带来了额外的认知负荷。     

How to present collision warnings at intersections?—A comparison of different approaches

Analyses of intersection accidents have shown that in most cases drivers overlook or see other road users with the right of way too late that they cannot react in time. Appropriate visual warning signals which support drivers in their attention allocation and driving behavior could be useful to improve drivers’ reaction. In a driving simulator study, two warning strategies varying in their timing: (1) top-down warning while approaching a critical intersection and (2) bottom-up warning directly before the critical incident were investigated in a T-intersection situation. For the bottom-up warning, two warning signal designs were compared. 48 subjects (M = 27.3 years, SD = 7.4 years) participated in the study. Driving data as well as subjective evaluation of the three warning signals (one early and two late warning signals) were analyzed. The early warning signal which was given while approaching the intersection showed a positive effect. Here, most collisions could be avoided due to drivers’ adaptation of their driving behavior toward safer driving. They waited longer at the intersection before turning and finally turned with a lower velocity. In addition, drivers evaluated the early warning signal as very “useful”. With regard to the late warnings a much smaller effect was found. From these results, requirements can be derived for the design of effective warning strategies when driving at critical intersections.     

A descriptive analysis of light vehicle-heavy vehicle interactions using in situ driving data

Two recently completed on-road in situ (naturalistic) data collection efforts provided a large data set in which to conduct an examination of crashes, near-crashes, and crash-relevant conflicts (referred to as critical incidents throughout this paper) that occurred between light vehicles (LV) and heavy vehicles (HV). Video and non-video data collected during the two studies were used to characterize critical incidents that were recorded between LV and HV drivers. Across both studies, 210 LV-HV critical incidents were recorded. Of these, 78% were initiated by LV drivers, while the remaining 22% were initiated by HV drivers. Aggressive driving, on the part of the LV driver, was found to be the primary Contributing Factor for LV driver-initiated incidents. For HV driver-initiated incidents, the primary Contributing Factor was poor driving techniques. These results suggest that future efforts at addressing LV-HV interaction incidents should include focusing on aggressive LV drivers. Additionally, it is recommended that HV drivers might benefit from improved driver training that includes instruction on defensive driving skills. The in situ methodology provides an alternative to traditional crash databases, developed from police accident reports, for studying crash causation and driver behavior.     

Mitigating driver distraction with retrospective and concurrent feedback

OBJECTIVES: An experiment was conducted to assess the effects of retrospective and combined retrospective and concurrent feedback on driver performance and engagement in distracting activities. BACKGROUND: A previous study conducted by the authors showed that concurrent (or real time) feedback can help drivers better modulate their distracting activities. However, research also shows that concurrent feedback can pose additional distractions due to the limited time and resources available during driving. Retrospective feedback, which is presented at the end of a trip (i.e., post-drive), can include additional information on safety critical situations during a trip and help the driver learn safe driving habits. METHOD: A driving simulator study was conducted with 48 participants and 3 conditions: retrospective feedback, combined feedback (both retrospective and concurrent), and no feedback (baseline case). RESULTS: The feedback conditions (retrospective and combined) resulted in faster response to lead vehicle braking events as depicted by shorter accelerator release times. Moreover, combined feedback also resulted in longer glances to the road. CONCLUSIONS: The results suggest that both feedback types have potential to improve immediate driving performance and driver engagement in distractions. APPLICATION: Combined feedback holds the most promise for mitigating the effects of distraction from in-vehicle information systems.     

Measurement of driver calibration and the impact of feedback on drivers’ estimates of performance

Recent studies focused on driver calibration show that drivers are often miscalibrated, either over confident or under confident, and the magnitude of this miscalibration changes under different conditions. Previous work has demonstrated behavioral and performance benefits of feedback, yet these studies have not explicitly examined the issue of calibration. The objective of this study was to examine driver calibration, i.e., the degree to which drivers are accurately aware of their performance, and determine whether feedback alters driver calibration. Twenty-four drivers completed a series of driving tasks (pace clocks, traffic light, speed maintenance, and traffic cones) on a test track. Drivers drove three different blocks around the test track: (1) baseline block, where no participants received feedback; (2) feedback block, where half of the participants received performance feedback while the other half received no feedback; (3) a no feedback block, where no participants received feedback. Results indicated that across two different calibration measures, drivers were sufficiently calibrated to the pace clocks, traffic light, and traffic cone tasks. Drivers were not accurately aware of their performance regarding speed maintenance, though receiving feedback on this task improved calibration. Proper and accurate measurements of driver calibration are needed before designing performance feedback to improve calibration as these feedback systems may not always yield the intended results.     

The impact of secondary task cognitive processing demand on driving performance

Crash causation research has identified inattention as a major source of driver error leading to crashes. The series of experiments presented herein investigate the characteristics of an in-vehicle information system (IVIS) task that could hinder driving performance due to uncertainty buildup and cognitive capture. Three on-road studies were performed that used instrumented passenger and tractor-trailer vehicles to obtain real-world driving performance data. Participants included young, middle-aged, and older passenger vehicle drivers and middle-aged and older commercial vehicle operators. While driving, they were presented with IVIS tasks with various information densities, decision-making elements, presentation formats, and presentation modalities (visual or auditory). The experiments showed that, for both presentation modalities, the presence of multiple decision-making elements in a task had a substantial negative impact on driving performance of both automobile drivers and truck drivers when compared to conventional tasks or tasks with only one decision-making element. The results from these experiments can be used to improve IVIS designs, allowing for potential IVIS task phenomena such as uncertainty buildup and cognitive capture to be avoided.     

Local traffic condition: improvement of a vehicle-based measurement approach

Advanced driver assistance systems (ADAS) can be designed to adapt to the driver?s current needs, for example, taking the traffic conditions in the immediate vicinity of the ego vehicle into account. This motivates the design of a system intended to measure the local traffic conditions in a way, which corresponds to the driver?s perception of traffic conditions. A new method of measurement is introduced. It is based on automotive sensor technology. A stochastic approach is employed to optimise and evaluate the relationship between the subjective measurements made by the driver and the objective measurements presented. To this end, the drivers? subjective ratings are taken as reference. The analysis presented here is based on simulation data. The outcomes show that the design of the developed simulation environment is applicable to this work. It is shown that the objective measures correspond to the driver?s subjective ratings. Thus, it is possible to estimate the driver?s perception of traffic conditions using a technical measurement. The algorithm in its presented form is appropriate to give an estimation, which can be used as an input parameter for ADAS.     

Microsimulation modelling of driver behaviour towards alternative warning devices at railway level crossings

Level crossings are amongst the most complex of road safety issues, due to the addition of rail infrastructure, trains and train operations. The differences in the operational characteristics of different warning devices together with varying crossing, traffic or/and train characteristics, cause different driver behaviour at crossings. This paper compares driver behaviour towards two novel warning devices (rumble strips and in-vehicle audio warning) with two conventional warning devices (flashing light and stop sign) at railway level crossings using microsimulation modelling. Two safety performance indicators directly related to collision risks, violation and time-to-collision, were adopted. Results indicated the active systems were more effective at reducing likely collisions compared to passive devices. With the combined application of driving simulation and traffic microsimulation modelling, traffic safety performance indicators for a level crossing can be estimated. From these, relative safety comparisons for the different traffic devices are derived, or even for absolute safety evaluation with proper calibration from field investigations.     

Experimental evaluation of fog warning system

Highway safety is a major concern to the public and to transportation professionals, so the number of crashes caused by poor visibility due to fog form an alarming statistic. Drivers respond to poor visibility conditions in different ways: some slow down; others do not. Many drivers simply follow the taillights of the vehicle ahead. Accordingly, hazardous conditions are created in which speeds are both too high for the prevailing conditions and highly variable. Findings are presented from a study of traffic crashes due to fog in the southern region of Saudi Arabia. The primary objective was to assess the effectiveness of fog detection and warning system on driver behavior regarding speed and headway. This warning system includes visibility sensors that automatically activate a variable message sign that posts an advisory speed when hazardous conditions due to fog occur. The system was installed on a 2 km section of a two-lane, rural highway. A data set of 36,013 observations from both experimental and control sections at two study sites was collected and analyzed. The data included vehicle speed, volume, and classification; time headway, time of day, and visibility distance. Although the warning system was ineffective in reducing speed variability, mean speed throughout the experimental sections was reduced by about 6.5 kph. This reduction indicates that the warning system appeared to have a positive effect on driver behavior in fog even though the observed mean speeds were still higher than the posted advisory speed. From relationships found in the literature between mean driving speed and number of crashes, a speed reduction of only 5 kph would yield a 15% decrease in the number of crashes.     

Overly cautious and dangerous: An empirical evidence of the older driver stereotypes

Negative comments regarding the competency of older drivers are frequently heard in the general population. While negative stereotypes of older drivers seem to be present, their existence has yet to be empirically validated. We thus investigated the stereotypes pertaining to older drivers in two experiments. In both experiments young adults viewed 12 simulated clips of three categories of driving behaviors (i.e., younger adults’ unsafe behaviors, older drivers’ unsafe behaviors and appropriate-safe driving behaviors) without knowing the driver's age. They were asked to rate how representative the behaviors were of a typical younger, middle-aged, or older driver. Experiment 1 showed that older drivers’ unsafe behaviors were rated as significantly more representative of the typical older driver, while young adults’ unsafe behaviors were perceived as significantly more representative of the typical younger driver. In Experiment 2, younger participants viewed the same clips but were only asked to indicate whether the observed behavior was representative or not of the typical older driver. The main findings were replicated. When asked to describe the main features of the typical older drivers, participants qualified them as being overly cautious, uncomfortable behind the wheel, and unsafe and dangerous. The potential implications on driving performance of older drivers and on driving cessation and are discussed.     

Perceptual and attentional effects on drivers' speed selection at curves

This paper describes an experiment comparing the relative effectiveness of various types of warnings on drivers' speed selection at curves. The experiment compared three types of curve warnings across three different curve types in a driving simulator. All of the warnings worked reasonably well for severe curves (45 km/h), regardless of demands from a secondary (cell phone) task. For less demanding curves, only those warnings with a strong perceptual component (i.e., implicit cues) were effective in reducing drivers' curve speeds in the presence of the cell phone task. The design implications of these data appear straightforward; curve warnings that contain perceptual components or emphasise the physical features of the curve work best, particularly in cognitively demanding situations. The cell phone task added to driver workload and drivers became less responsive to primary task demands (i.e., speeds were elevated and reaction times were longer).