Driver distraction based lane-keeping assistance

Driver assistance systems could be much more effective, if they were adaptive to the driver’s state. Following the idea of a lane-keeping assistance system, which is adaptive to the use of in-vehicle information systems (IVIS), two field experiments were conducted. The first experiment concerned the influence of typical IVIS tasks on lane-keeping. Most IVIS tasks increased the lateral deviations, but in the majority of cases the drivers were still able to stay in their lane.The second experiment concerned two questions: (1) how much lateral support would be needed and (2) how drivers accept this kind of support. The results show that all lateral support algorithms increased the lane-keeping performance with the algorithms providing a higher amount of assistance proving the most useful. All assistance systems were rated as helpful and were considered to increase driver safety, both by the drivers who did not have problems in lane-keeping without assistance.In light of these results an adaptation of a lateral support system to the IVIS-use seems to be useful and worthwhile.     

Should I Drive Today? Development of a daily driving decisions model for older adults

BackgroundOlder adults in communities make daily decisions about how to meet their transportation needs so they can access services and stay socially connected. With the aging of populations in developed countries, the travel decisions of older adults will have increasing impacts. Research studies have identified different sets of factors that contribute to certain travel decisions, but little research has been directed towards understanding how individuals select information from all available factors, what information they include in their decisions under different circumstances, and the processes they use in making their transportation decisions.MethodsThis exploratory study involved 20 men and 17 women, mean age 78.6 years (range 70–96), who drove weekly. All participants were involved in each phase of the 3-phase study. In Phase 1, a review of the literature and interviews with the participants was used to collect information, and inductive thematic analysis was employed to construct a draft conceptual model of older driver decision-making. In Phase 2, participants completed a stated preference task of written scenarios to demonstrate their decision-making strategies. Results were tabulated and used to refine a final Daily Driving Decisions model. In Phase 3, a card sorting decision task was used to test the model with participants.ResultsThe final dynamic Daily Driving Decisions Model was confirmed to describe decision processes used by the participants in making decisions about how they would meet their transportation needs. The model describes three categories of factors used in decisions, labelled Motivators, Constraint/Enablers and Context, each containing four attribute themes. A significant finding was the variable use of the same item to either constrain or enable the decision to drive depending on the variation of other factors in the scenario. Participants demonstrated use of compensatory and noncompensatory (heuristic, habitual) decision processes that were accommodated by the model.ConclusionThe proposed Daily Driving Decisions Model addresses a gap in our understanding of how older drivers make their decisions about meeting their transportation needs. The model presents a template for classifying the types of information used, ignored or discarded by older adults, and the pathways that they take to arrive at their decisions. The model provides opportunities for further research in testing the influence of other factors such as urban/rural residence, income, health status and culture on driving decisions. Further, the model can be used by practitioners to gain insight into the decision-making behaviours of individuals and to develop interventions to enhance their decision-making skills.     

The influence of road familiarity on distracted driving activities and driving operation using naturalistic driving study data

This paper analyzed the influence of familiarity on the involvement of secondary tasks and driving operation using naturalistic driving study (NDS) data. Distracted driving activities were extracted from face videos captured in 557 trips, including 501 trips on familiar roads and 56 trips on unfamiliar roads. These trips were completed by 155 drivers using their own vehicles during daytime hours under good weather conditions. The data showed the frequency of distracted driving activities and duration time were higher on familiar roads compared to unfamiliar roads. More types of secondary tasks were found on familiar roads. Focusing on objects was the most common distracted driving activity on familiar roads. The average time drivers used to eat or drink was highest (8.67 s) on familiar roads. The time drivers spent checking their cell phone was high on both familiar roads and unfamiliar roads. Since driving operation is directly related to crash risk, this paper also analyzed the difference of driving operation on familiar roads and unfamiliar roads. The speed profiles were generated on well-known versus unfamiliar roads. It was shown that drivers were more likely to be speeding and select a short distance to deceleration near the intersections. The findings indicated that distracted driving phenomenon was more serious on familiar roads.     

Modelling the human lane-change execution behaviour through Multilayer Perceptrons and Convolutional Neural Networks

Driving is a highly complex task that involves the execution of multiple cognitive tasks belonging to different levels of abstraction. Traffic emerges from the interaction of a big number of agents implementing those behaviours, but until recent years, modelling it by the interaction of these agents in the so called micro-simulators was a nearly impossible task as their number grows. However, with the growing computing power it is possible to model increasingly large quantities of individual vehicles according to their individual behaviours. These models are usually composed of two sub-models for two well-defined tasks: car-following and lane-change. In the case of lane-change the literature proposes many different models, but few of them use Computational Intelligence (CI) techniques, and much less use personalization for reaching individual granularity. This study explores one of the two aspects of the lane-change called lane-change acceptance, where the driver performs or not a lane-change given his intention and the vehicle environment. We demonstrate how the lane-change acceptance of a specific driver can be learned from his lane change intention and surrounding environment in an urban scenario using CI techniques such as feed-forward Artificial Neural Network (ANN). We work with Multilayer Perceptron (MLP) and Convolutional Neural Networks (CNN) architectures. How they perform one against the other and how the different topologies affect both to the generalization of the problem and the learning process are studied.     

Adaptive warning signals adjusted to driver passenger conversation: Impact of system awareness on behavioral adaptations

The study investigated behavioral adaptation caused by warning signals that adaptively support drivers engaged in a passenger conversation. Novel advanced driver assistance systems (ADAS) could monitor the driver state to provide warnings adjusted to the driver’s current need for support. While first research indicates positive effects of dynamically adjusting ADAS, the overall safety potential of such adaptive systems remains unclear as little is known about adverse behavioral effects. Occasional inappropriate support due to an incorrect prediction of the current driver state might lead to critical situations when drivers are aware of the adaptive nature of the system and thus rely on the expected system behavior. To better understand behavioral adaptation to dynamically adjusting warnings, 46 participants driving on a test track reacted to two types of warnings (auditory, vibrotactile-auditory) while engaging in driver-passenger conversation or not. In a compensating warning strategy, vibrotactile-auditory warnings were displayed during conversations and auditory warnings in situations without conversation. This strategy was compared to a non-compensating warning strategy in which these assignments were reversed. The impact of behavioral adaptation was measured by considering reactions to simulated failures of these strategies. The role of system awareness for behavioral adaptation was investigated by manipulating awareness of these warning strategies across groups. We found that vibrotactile-auditory warnings reduced the detrimental effects of conversation on reaction times. Crucially, adverse behavioral adaptation was observed whenever an expected vibrotactile-auditory warning was replaced by an auditory warning but this effect was restricted to drivers in the awareness group. The results show that adaptive warning signals optimize the effectiveness of warnings during driver-passenger conversation but adverse behavioral adaptation develops when drivers are aware of the underlying warning strategy. This implies that future adaptive systems are less likely to be associated with behavioral adaptation if the adaptive nature of the system remains unnoticeable to the driver. Our findings could be used for developing and optimizing user-centered ADAS.     

“What Makes a Cooperative Driver?” Identifying parameters of implicit and explicit forms of communication in a lane change scenario

An automated vehicle needs to learn how human road users communicate with each other in order to avoid misunderstandings and prevent giving a negative outward image during interactions. The aim of the present work is to develop an autonomous driving system which communicates its intentions to change lanes based on implicit and explicit rules used by human drivers. To reach this goal, we aimed at gaining a deeper understanding of which aspects of lane change behaviour makes them cooperative from the perspective of other drivers. Therefore a vehicle used various lane change announcement strategies by varying combinations of driving parameters in a static driving simulator. (First study: Start indicator signal, Waittime, lane change duration; Second study: Longitudinal acceleration). It’s impact on the perception and behaviour of other road users was observed in two studies (N = 25 per study). The results showed that the earlier the merging vehicle was indicating its intentions, the more cooperative it was perceived. When turning on the indicator at a later time participants considered it as more cooperative to merge with a slower or faster lane change duration or to wait longer in the lane before starting to move to the other lane. An early longitudinal acceleration when starting to change lanes is perceived more cooperative. These findings can be used to model a lane change strategy based on human behaviour, which will eventually lead to more acceptable and safer interactions between automated and non-automated road users.     

Cross-cultural comparison of drivers’ tendency to commit different aberrant driving behaviours

The first aim of the present study was to identify key items which are rated differently by drivers from Finland, Sweden, Greece and Turkey. The second aim was to examine how these key items relate to drivers’ self-reported accident involvement. Similar comparisons have previously been conducted in Europe but these have only included items classified as violations and errors, but not lapses. A sample of Finnish (N = 200), Swedish (N = 200), Greek (N = 200) and Turkish (N = 200) drivers completed the driver behaviour questionnaire (DBQ) and reported their accident involvement during the previous 3 years. The results showed that nine key items (which drivers from different countries rated differently) could be identified. These items included two aggressive violations, four ordinary violations, three lapses, but no errors. Out of these nine items, five items (Become angered by a certain type of driver and indicate your hostility by whatever means you can, Disregard the speed limit on a motorway, Overtake a slow driver on the inside, Pull out of a junction so far that the driver with right of way has to stop and let you out and Get into the wrong lane approaching a roundabout or a junction) could explain differences in drivers’ self-reported yearly accident involvement when all four countries were taken together. At the same time, none of the items could explain differences in self-reported yearly accident involvement in Finland and Sweden while one of the items (Overtake a slow driver on the inside) could explain differences in self-reported yearly accident involvement in Greece and two of the items (Become angered by a certain type of driver and indicate your hostility by whatever means you can and Disregard the speed limit on a residential road) could explain differences in self-reported yearly accident involvement in Turkey. This shows that different countries have different problems with regard to aberrant driving behaviours which need to be taken into account when promoting traffic safety interventions and the driver behaviour questionnaire (DBQ) can be used to diagnose risk areas and to better inform road safety practitioners within and between countries.     

How drivers adapt their behaviour to changes in task complexity: The role of secondary task demands and road environment factors

The present study investigates the impact of different sources of task complexity such as driving demands and secondary task demands on driver behaviour. Although much research has been dedicated to understanding the impact of secondary task demands or specific road traffic environments on driving performance, there is little information on how drivers adapt their behaviour to their combined presence. This paper aims to describe driver behaviour while negotiating different sources of task complexity, including mobile phone use while driving (i.e., calling and texting) and different road environments (i.e., straight segments, curves, hills, tunnels, and curves on hills). A driving simulator experiment was conducted to explore the effects of different road scenarios and different types of distraction while driving. The collected data was used to estimate driving behaviour through a Generalized Linear Mixed Model (GLMM) with repeated measures. The analysis was divided into two phases. Phase one aimed to evaluate driver performance under the presence and absence of pedestrians and oncoming traffic, different lanes width and different types of distraction. The second phase analysed driver behaviour when driving through different road geometries and lane widths and under different types of distraction. The results of the experiment indicated that drivers are likely to overcorrect position in the vehicle lane in the presence of pedestrians and oncoming traffic. The effect of road geometry on driver behaviour was found to be greater than the effect of mobile phone distraction. Curved roads and hills were found to influence preferred speeds and lateral position the most. The results of this investigation also show that drivers under visual-manual distraction had a higher standard deviation of speed and lateral position compared to the cognitive distraction and the non-distraction condition.     

Safety effects of parking maneuvers

Parking maneuvers, particularly a vehicle’s maneuver for entering and leaving a parking space, have varying rates of use and safety impacts. In this effort, crash data were collected for parking lots in the vicinity of a university campus and compared to observational parking position data. The campus was selected for this study because a change in the parking enforcement process was expected to (and did) change parking maneuver choices. When entering and leaving a parking space, three maneuver options exist for drivers: (1) forward, (2) reverse, and (3) pulling through an adjacent parking space. When specifically entering a parking space, the maneuver options become: (1) pull-in, (2) back-in, and (3) pull-through. When leaving the parking space, the maneuver options become: (1) pull-out, (2) back-out, and (3) pull-through. This study found that the pull-in/back-out vehicle maneuver’s percentage of total crashes was greater than the percentage of vehicles that were actually observed to use the same maneuver. The analysis from this study implies that the pull-in/back-out parking maneuver is more likely to result in a collision and therefore, is associated with a higher crash risk. Further analysis of North Carolina’s parking related fatal and serious injury crashes found that vehicles backing out of parking spaces was overwhelmingly the main cause for these serious injuries. 90% of North Carolina’s parking related fatal and serious injuries occurred during a back-out maneuver. Overall, this study concludes that the back-in/pull-out parking maneuver is safer than the pull-in/back-out maneuver and is the recommended approach to 90° parking.     

Evidence for the role of personality in the cognitive performance of older male drivers

ObjectivesThis study aimed to examine whether in older male drivers the level of information processing under increasing cognitive workload varies with the level of personality traits.MethodThe study involved 60 male, active drivers aged between 65 and 81 (M = 71, SD = 4.4). During passive driving they performed central (a test in the n-back scheme, in three cognitive workload levels) and peripheral (memorizing objects characteristic for road traffic) tasks presented simultaneously in the central and peripheral visual fields. A wide-angle high-fidelity collimated visual system was used to display visual stimuli. In the central task, the reaction time and its correctness were measured, while the peripheral task measured the number of memorized objects. The EPQ-R questionnaire was used to assess personality traits, and the memory functions were assessed using the MMSE test.ResultsAs the central task-induced cognitive workload increases, the performance of both central and peripheral tasks deteriorates. These tasks were performed less accurately by drivers with a higher level of psychoticism than those with a lower level of this trait. This dependency related to all levels of cognitive workload. In contrast, the drivers with a higher level of neuroticism showed a lower level of performance of the peripheral task than those with a low level of this trait. This dependency concerned only the central task with a medium cognitive workload.ConclusionThe level of personality traits, especially the ones related to the emotional sphere, contributed to the decrease in the effectiveness of action (longer reaction time, increase in the number of errors, decrease in the number of memorized objects) in the conditions of increasing cognitive workload. A higher level of neuroticism and psychoticism can be conducive to making errors on the road. This applies especially to situations that pose high demands on the driver’s cognitive system.