Research Directions

———————————————–

Automated Driving

Drivers’ anticipation behaviours in autonomous vehicles

Despite the ambitious plans of vehicle manufacturers, vehicle technologies usually take two to five decades to saturate their potential market, and currently, SAE level-2 automation (SAE J3016_201401) is the state-of-the-art vehicle-automation technology. For this level, drivers are required to visually monitor the driving environment and the automation and anticipate how a given situation may evolve, in order to intervene in a timely manner when necessary. A good design needs to support drivers’ anticipation of both the environment as well as the automated system that they are interacting with. Anticipation in driving in a non-automated context has been defined as “a manifestation of a high-level cognitive competence that describes the identification of stereotypical traffic situations on a tactical level through the perception of characteristic cues, and thereby allows for the efficient positioning of a vehicle for probable, upcoming changes in traffic.” (Stahl, Donmez, & Jamieson, 2014, p. 603). Anticipatory driving has been shown to be more prevalent among experienced drivers (Stahl et al., 2014), but to be inhibited by engagement in secondary activities (He & Donmez, 2018). Anticipatory driving, although arguably an important skill for supervising level-2 vehicles, has not yet been studied for automated vehicles.  This project aims to investigate (1) how automation affects anticipation in driving and (2) what type of feedback can be effective to support driver anticipation in automated vehicles.

Sponsor: NSERC (Discovery Grant)

Student PI(s): Dengbo He

Related publication(s):

Training to improve drivers’ mental models of ADAS and support appropriate reliance

Advanced driver assistance systems (ADAS) are available in an increasing number of vehicles, however, drivers receive little or no training on how to use ADAS safely. This research focuses on studying drivers’ mental models of ADAS, specifically adaptive cruise control (ACC) and lane keeping assist (LKA), which is important to support them in using the automation safely. Through a survey, we found that many drivers overestimate the capabilities of ACC and LKA and that drivers who owned vehicles with ACC and LKA did not seem to have a better understanding of these systems compared to drivers who had never used ACC or LKA. Participants with a worse understanding of ACC and LKA had higher trust in these systems, and higher trust was associated with a greater self-reported likelihood of engaging in secondary tasks while driving. Future studies will involve developing different training methods and investigating their effectiveness at improving drivers’ understanding of ACC and LKA and supporting appropriate trust and reliance on these systems.

Sponsor: NSERC (Discovery Grant)

Student PI(s): Chelsea DeGuzman

Driver Distraction

Naturalistic driving

Distracted driving is a significant concern for traffic safety. We use the SHRP2 NEST dataset to investigate distraction engagement while driving. Our findings reveal that drivers modulate their distraction engagement based on environmental demands and the chosen speed. Environments with higher visual difficulty (e.g., busy urban areas during rainy weather) decrease the likelihood of distraction engagement, whereas speed increase is associated with a decrease in the likelihood of distraction engagement only in higher motor control difficulty situations (i.e., curvy and wet roads). Our research has been featured in Toyota CSRC Report for 2017.

Sponsor(s): Toyota Collaborative Safety Research Center (CSRC)

Student PI(s)

Related Publication(s): 

  • Risteska, M., (2018). Exploration of naturalistic driving data: Development of distracted driver behaviour models (MASc Thesis). University of Toronto.
    [Link]
  • Donmez, B., Chen, H. Y. W., & Risteska, M. (2017). Naturalistic Engagement in Secondary Tasks (NEST): Driver Behavior and Secondary Task Engagement in Crashes and Near-Crashes. (HFASt2017-01), University of Toronto Human Factors and Applied Statistics Laboratory. Technical report submitted to Toyota Collaborative Safety Research Center.
    [Link]
  • Domoyer, J. E., Lee, J., Reimer, B., Seaman, S., Angell, L., Zhang, C., & Donmez, B. (2016). SHRP2 NEST Database: Exploring conditions of secondary task engagement in naturalistic trip data. In Proceedings of the 8th International Conference on Automotive User Interfaces and Interactive Vehicular Applications, Ann Arbor, MI.
    [Link]

Effects of searching for street parking on driver behaviour

Searching for street parking adds to traffic congestion and time/fuel wasted for drivers however its effect on driving behaviour and driver performance is unknown. This study investigates how searching for street parking affects driver physiology, performance, and visual attention allocation. Participants drive a vehicle instrumented with sensors and cameras in a busy urban centre (downtown Toronto). They will also be outfitted with physiological sensors (ECG and GSR) and a head-mounted eye tracker. The analysis of the physiological and vehicle data is expected to draw conclusions on increased mental workload of drivers while searching for street parking as well as the effect on their driving performance.

Sponsor(s): NSERC (Discovery Grant)

Student PI(s): Canmanie Teresa Ponnambalam; George Liu

Emerging technologies and driver distraction

Smartwatches and other wearables are new devices that are making their way into the driving space, and have been marketed as being easily accessible and glanceable ways of getting information. We conducted two driving simulator experiments comparing how drivers interact and respond to notifications using smartwatches and smartphones. We found that our participants spent more time looking at smartwatches than smartphones when they read notifications. Our participants were also slower in reaction to a lead vehicle braking while responding to notifications on the smartwatch than the smartphone. These results show that smartwatches may have detrimental effects on driving that are similar to or worse than smartphones, and can be a potential source of distraction on the road.

Sponsor(s): NSERC (Discovery Grant)

Student PI(s): Wayne Giang

Related Publication(s):

  • Giang, W. C. W., Chen, H. Y. W., & Donmez, B. (2017). Smartwatches vs. smartphones: Notification engagement while driving. International Journal of Mobile Human Computer Interaction, 9(2), 40-58.
    [Link]
  • Giang, W. C. W., Shanti, I., Chen, H. Y. W., Zhou, A., & Donmez, B. (2015). Smartwatches vs. smartphones: A preliminary report of driver behavior and perceived risk while responding to notifications. In Proceedings of the 7th International Conference on Automotive User Interfaces and Interactive Vehicular Applications, Nottingham, UK.
    [Link]
  • Giang, W. C. W., Hoekstra-Atwood, L., & Donmez, B. (2014). Driver engagement in notifications: A comparison of visual manual interaction between smartwatches and smartphones. In Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting, Chicago, IL.
    [Link]

Perception-response time to emergency roadway hazards and the effect of cognitive distraction

A driver’s ability to detect potential hazards and respond accordingly is crucial for traffic safety. The main objectives of this study are to (1) investigate the use of eye movement recordings combined with other motor response measurements in order to sub-divide a driver’s perception-response time interval when responding to an emergency roadway hazard, (2) to use this method of sub-diving to gain further insight into the effects of cognitive distraction on a driver’s ability to perceive and respond to an emergency roadway hazard by analyzing the effects at each stage.

Sponsor(s): NSERC (Discovery Grant)

Student PI(s)

Effects of driver distraction on crash injury severity and crash type

The objective of this research is to determine how different distraction types interact with driver age to influence crash outcomes. The particular crash outcomes we focus on consist of the crash type as well as the injury severity sustained by occupants involved in the crash. The findings of this research will have implications for policy making as well as prioritizing capabilities of distraction-related safety systems.

Sponsor(s): NSERC (Discovery Grant)

Student PI(s):

Effects of distractions on injury severity in police-involved crashes

A police cruiser can have multiple devices integrated in the cab, such as a laptop, radio, as well as strobe and siren controls. Although distractions might be a concern for police drivers, the effects of distractions on police-involved crashes have not been empirically studied before. As a first step in addressing this research gap, we built an ordered logit model to investigate the likelihood of severe injuries when a crash involves distracted police drivers.

Sponsor(s): University of Toronto (Start-up grant)

Student PI(s):

Volunrable Road Users’ Safety

Instrumented vehicle studies examining driver attention towards vulnerable road users at intersections

The role of human factors research is to provide an understanding of how drivers perform as a system component in the safe operation of vehicles. This role recognizes that driver performance is influenced by many environmental, psychological, and vehicle design factors. An on-road study may provide such high level of ecological validity. Crash data indicate that misallocation of attention is a major source of conflicts particularly with vulnerable road users (pedestrians and cyclists) at intersections. A better understanding of how and why attentional failures occur can inform the design of interventions (e.g., educational programs for high-risk drivers, infrastructure design) to enhance overall vulnerable road user safety.

It is known that drivers experience increased visual and mental demands while driving through or making turns at intersections, as intersections require drivers to divide their attention in several directions and toward a variety of traffic participants (e.g., other vehicles, pedestrians, cyclists) and control devices (e.g. road signs, traffic signals). This project’s fundamental aim is to enhance vulnerable road user safety at intersections by collecting data through a state-of-the-art instrumented vehicle and cutting-edge eye-tracking equipment. For this purpose, each participant was instructed to drive through the pre-determined routes in downtown Toronto including various types of intersections in order to capture the effects of road design on attentional failures. To assess individual differences, each participant was asked to complete surveys (e.g., Driver Behavior Questionnaire, Cognitive Failures Questionnaire, Arnett Inventory of Sensation Seeking, NASA TLX) as well as perform computerized attention tasks examining general attention abilities in a laboratory setting.

Student PI(s): Nazli Eser Kaya

Driver Feedback

Changing driver behaviour through social norms

Distraction is a significant concern for teen drivers, because they more widely use technologies that are key sources of driver distraction and they lack critical strategic and tactical skills possessed by more experienced adult drivers. Parents and peers significantly impact the values, beliefs, and behaviours of teenagers. Through a survey study, we found that teenagers misperceived their parents’ and peers’ level of engagement and approval of distracted driving. Providing distraction-related feedback to teenagers on what their parents and peers do and think can help mitigate this dangerous activity by correcting such misperceptions. For example, such a system can show the teens how they compare to their parents in terms of engagement in distracting activities and the resulting driving performance. In the simulator, we showed that a social norms feedback system based on what the parents do can reduce the level of distraction engagement among the teens.

Sponsor(s): Toyota CSRC, AUTO21

Student PI(s): Maryam Merrikhpour

Related Publication(s): 

  • Merrikhpour, M. & Donmez, B. (2017). Designing feedback to mitigate teen distracted driving: A social norms approach. Accident Analysis and Prevention, 104, 185-194.
    [Link]
  • Merrikhpour, M. & Donmez, B. (2016). Social norms and teenage driver distractions. In Proceedings of the 26th Canadian Association of Road Safety Professionals Conference, Halifax, Nova Scotia. Student Paper Competition, Honorable Mention.
    [Link]
  • Merrikhpour, M. & Donmez, B. (2016). Designing feedback to mitigate teen distracted driving behaviour: A social norms approach. (HFASt2016-02), University of Toronto Human Factors and Applied Statistics Laboratory. Technical report submitted to Toyota Collaborative Safety Research Center.
    [Link]
  • Merrikhpour, M. & Donmez, B. (2015). Interim Report. Designing feedback to mitigate teen distracted driving behaviour: A social norms approach. (HFASt2015-03), University of Toronto Human Factors and Applied Statistics Laboratory. Technical report submitted to Toyota Collaborative Safety Research Center.
    [Link]

Changing young drivers’ behaviours using gamification

Introducing elements commonly found in games – such as points and leaderboards – can increase the motivation towards and enjoyment of engaging in positive behavioral changes. We designed a gamified in-vehicle system dubbed RoadHero, that incorporates an overarching theme, avatars, and badges, to encourage young drivers to adopt safer driving behaviors. In a simulator study, we found that this gamified system reduced the rate and duration of long glances away from the road, which are known to increase crash risks. These initial findings show that gamification can reduce distracted driving in younger drivers, although much remains to be explored in applying gamification to road safety. Our research has been featured in Toyota CSRC Report for 2017.

Sponsor(s): Toyota CSRC, AUTO21

Related Publication(s):

  • Xie, J., Chen, H. Y. W., & Donmez, B. (2016). Gaming to safety: Feedback gamification for mitigating driver distraction. In Proceedings of the Human Factors and Ergonomics Society 60th Annual Meeting (pp. 1884-1888), Washington, DC. Surface Transportation Technical Group Best Student Paper Award.
    [Link]
  • Xie, J. (2016). Gaming to safety: The design and evaluation of feedback gamification for mitigating driver distraction (MASc Thesis). University of Toronto.
    [Link]
  • Xie, J., Chen, H. Y. W., & Donmez, B. (2015). Designing feedback to induce safer driving behaviors: The influence of feedback timing and motivation on driving performance and distraction engagement. (HFASt2015-02), University of Toronto Human Factors and Applied Statistics Laboratory. Technical report submitted to Toyota Collaborative Safety Research Center.
    [Link]

Student PI(s): Jeanne Xie

Modifying driver behavior using a reward system 

Providing feedback and rewards to drivers may help them reduce their dangerous driving habits. We looked at the effects of a feedback-reward system on speeding and tailgating behaviours. Data was collected from 37 drivers in a field trial where feedback and rewards were provided to the drivers based on complying with speed limits and maintaining safe headway distances. Drivers collected rewards for compliant behaviour. In addition, an in-vehicle display provided drivers with real-time feedback when they were driving, in the form of warnings when they exceeded the speed limit or when they were tailgating. The system resulted increased speed limit and headway compliance. When the system was removed, the positive effect for speed limit compliance continued.

Sponsor(s): Connaught New Researcher Award, Transport Canada

Student PI(s):

Related publication(s)

  • Merrikhpour, M., Donmez, B., & Battista, V. (2014). A field operational trial evaluating a feedback-reward system on speeding and tailgating behaviors. Transportation Research Part F: Psychology and Behaviour, 27, 56-68.
    [Link]
  • Merrikhpour, M. (2013). Effects of a feedback-reward system on speeding and tailgating behaviours (MASc Thesis). University of Toronto.
    [Link]
  • Merrikhpour, M., Donmez, B., & Battista, V. (2012). Effects of a feedback/reward system on headway maintenance. In Proceedings of the Canadian Multidisciplinary Road Safety Conference, Banff, AB.
    [Link]
  • Merrikhpour, M., Donmez, B., & Battista, V. (2012). Effects of a feedback/reward system on speed compliance rates and the degree of speeding during noncompliance. In Proceedings of the Transportation Research Board 91st Annual Meeting, Washington, D.C
    [Link]

Facilitating anticipatory driving strategies

The project aims to aid in the development of conscious and competent driving skills, specifically with respect to drivers’ abilities to anticipate upcoming events in traffic. By enabling drivers to correctly interpret the current traffic situation and predict probable changes a few seconds ahead, adequate action can be taken early, and positively impact safety and fuel consumption.

The project is characterized by two main objectives, one being the identification of a working definition of anticipatory driving (and in this context, creation of a taxonomy of anticipatory driving and identification of appropriate measures), and the second relating to researching adequate means of supporting anticipatory driving. Our vision here is an in-car interface that provides an augmented, real-time representation of the current traffic situation, and facilitates adequate skill- and rule-based behaviours for anticipatory driving.

In conjunction with Prof. Greg A. Jamieson (University of Toronto)

Sponsor(s): AUTO21

Student PI(s): Patrick Stahl

Modifying driver behaviour using feedback

The primary objective of this project is to identify the relationship between driver attitudes towards safe driving and driver behaviour in response to feedback. This objective will be achieved by performing statistical analysis on a dataset collected through an on-road experiment which was conducted by Transport Canada. The findings of this research will have implications for the design of distraction-related safety systems.

Sponsor(s): Connaught New Researcher Award, Transport Canada

Student PI(s)

In-Vehicle Technology

Estimating drowsiness in drivers through wearable devices

The driver state can be monitored using vehicle-based (e.g., speed, acceleration), physiological (e.g., heart rate, brain activity), and face and body expression (e.g., head position, eye movements) measures (Aghaei et al., 2016). Although there has been extensive research on the application of driver state monitoring techniques in non-automated driving, most of the previous research used research-grade devices, which are not suitable or too expensive for in-vehicle application. Recent advancement in technologies allows for integration of a variety of sensors into smaller, more affordable and less intrusive wearable (e.g., smartwatch) and portable (e.g., smartphone) devices. For example, recent models of Apple Watch and Fitbit are equipped with sensors to monitor heart rate or even ECG; Garmin watch provides a function of measuring blood oxygen saturation level; Muse, a consumer-based headband, can record users’ EEG from channels on the forehead, along with users’ heart rate (HR), head movement and respiration in real-time. The increasing computing power in smartphones also makes it possible to assess driver state and even use cameras to analyze drivers’ gaze location and facial expressions in real-time. All these technologies provide the potential for low-cost in-vehicle driver state monitoring.

The objectives of this project are thus to design experiments and collect data to test the feasibility of different measures in identifying drowsiness in drivers, and train machine learning models using the collected data to estimate driver drowsiness.

Student PI(s): Dengbo He; Ziquan (William) Wang

Combining vehicle, physiological, visual data to provide driver feedback

This project investigates how mental distractions could influence drivers’ physiological states, facial expressions, and driving performance and further explores how physiological states, facial expressions and driving performance can be used to estimate drivers’ states. The cognitive state of drivers at varying levels of mental workload will be examined using the following physiological measures: heart rate, galvanic skin response (GSR), respiration and electroencephalogram (EEG), as well as information on facial expression and driving performance information. We also adopted and modified the commonly used n-back task to make it suitable for imposing cognitive load in in-vehicle EEG-related studies. Ultimately, this project aims to transform potentially distracting technologies into adaptive interfaces that offer constructive feedback to enhance driving experience by maintaining an optimal level of driver arousal during various driving conditions. 

Student PI(s): Dengbo He

Related Publication(s): 

  • He, D., Liu C., Donmez, B., & Plataniotis, K. (2017). Assessing high cognitive load in drivers through Electroencephalography. In Proceedings of the Transportation Research Board 96th Annual Meeting (17-02615), Washington, D.C.
    [Link]
  • Aghaei, A.S., Donmez, B., Liu, C.C., He, D., Liu, G., Plataniotis, K.N., Chen, H.Y.W., & Sojoudi, Z. (2016). Smart driver monitoring: When signal processing meets human factors. IEEE Signal Processing Magazine, 33(6), 35-48.
    [Link]

Enhancing voice-activation technologies to improve driving behaviour

Voice-activated infotainment systems are becoming increasingly common in modern vehicles, especially as an alternative to the more dangerous and often illegal option of hand-held cell phone usage. However, the voice-activated systems are not without their shortcomings. In the presence of loud background noise, such as talking passengers or music, voice recognition accuracy decreases. Working with an industry leader in voice control technologies, this project aims to evaluate how drivers interact with and feel about a novel voice control system which can still work effectively under heavy background noise.

Student PI(s): Neil Sokol; Joey Chakraborty

Related Publication(s):

  • Sokol, N., Chen, H. Y. W., & Donmez, B. (2017). Voice-controlled in-vehicle systems: Effects of voice-recognition accuracy in the presence of background noise. In Proceedings of the 9th International Driving Symposium on Human Factors in Driver Assessment, Training, and Vehicle Design, Manchester Village, VT.
    [Link]

Identifying risky driving styles through a GPS-enabled telematics platform

This research will identify if correlations exist between crash/moving violation histories and driving styles as measured through steering, speed and acceleration profiles obtained in a naturalistic on-road study.

Sponsor(s): FedDev (ARC), Skymeter Corp

Student PI(s):

Distractions in Operating Rooms

Systematic investigation of positive interruptions in healthcare

The objectives of this project are to examine and differentiate between positive and negative interruptions in a healthcare environment. Interruptions were studied extensively in the past but with a focus on their negative effects. Although many types of interruptions result in a break-in-task, in some cases interruptions communicate important information associated with patient’s safety. The majority of previous interruption research use a reductionist approach to minimize or prevent interruptions, and minimal attention has been given to the differentiation between positive and negative interruptions. This approach to the study of interruptions is a novel one and will have implications for how interruptions are approached in healthcare settings.

Sponsor(s): NSERC (PGS D)

Student PI(s):

Emergency Medical Transport

Understanding and supporting human evidence based decision making for emergency medical transport

The objective of this project is to design a decision support tool to aid emergency transport dispatch in selecting modes of transportation. Historical data on both air and land ambulance operations will be utilized to predict patient time to definite care (i.e., delivery of patient to receiving hospital). An interface will be designed which will provide the information needed by dispatch operators to make time-critical decisions under uncertainty. Project in conjunction with Dr. Russell MacDonald (ORNGE, University of Toronto).

Sponsor(s): ORNGE, NSERC (PGS D)

Student PI(s): Wayne Giang

Related Publication(s): 

Others

Mine traffic optimization

The project aims at optimizing traffic in open pit mines, and is a collaborative endeavour between University of Toronto, Queen’s University, and Barrick Gold Corporation. Two topics that are of particular interest to the University of Toronto team include the interactions between the haul truck operator and the partially automated dispatch system, as well as haul truck operator fatigue and distraction.Project in conjunction with Profs. Greg Jamieson (University of Toronto), Laeeque Daneshmend (Queen’s University), Josh Marshall (Queen’s University).

Sponsor(s): Barrick Gold, MITACS (Accelerate Internship Cluster)

Student PI(s):

Usability, pilot testing, and evaluation of B*Focused system

B*Focused is a system that is designed by Kangaroo Design to increase productivity in collaborative work environments. This project will evaluate B*Focused system.

In conjunction with Dr. Beth Savan (Sustainability Office, University of Toronto).

Sponsor(s): FedDev (ARC), Kangaroo Design

Student PI(s):