![The autonomous car is communicating with pedestrians at a crosswalk indicating that it is safe to cross. Source: [8].](https://www.researchgate.net/profile/Amir-Rasouli/publication/325464281/figure/fig1/AS:632192610865152@1527737872475/The-autonomous-car-is-communicating-with-pedestrians-at-a-crosswalk-indicating-that-it-is_Q320.jpg)
![Different concepts of communication for autonomous vehicles. a) Mercedes-Benz zebra crossing projection [19], b) Mitsubishi forward indicator [115], c) LEDs indicating yield [18], d) AutonoMI pedestrian detection and tracking indicator [111], e) advisory display [99], and f) AEVITA moving eye concept [116] (source [46]).](https://www.researchgate.net/profile/Amir-Rasouli/publication/325464281/figure/fig5/AS:632192615075848@1527737873418/Different-concepts-of-communication-for-autonomous-vehicles-a-Mercedes-Benz-zebra_Q320.jpg)
Content uploaded by Amir Rasouli
Author content
All content in this area was uploaded by Amir Rasouli on Jun 02, 2018
Content may be subject to copyright.
A preview of the PDF is not available
Autonomous Vehicles that Interact with Pedestrians:A Survey of Theory and Practice
Preprints and early-stage research may not have been peer reviewed yet.
Abstract and Figures
One of the major challenges that autonomous cars are facing today is driving in urban environments. To make it a reality, autonomous vehicles require the ability to communicate with other road users and understand their intentions. Such interactions are essential between the vehicles and pedestrians as the most vulnerable road users. Understanding pedestrian behavior, however, is not intuitive and depends on various factors such as demographics of the pedestrians, traffic dynamics, environmental conditions, etc. In this paper, we identify these factors by surveying pedestrian behavior studies, both the classical works on pedestrian-driver interaction and the modern ones that involve autonomous vehicles. To this end, we will discuss various methods of studying pedestrian behavior, and analyze how the factors identified in the literature are interrelated. We will also review the practical applications aimed at solving the interaction problem including design approaches for autonomous vehicles that communicate with pedestrians and visual perception and reasoning algorithms tailored to understanding pedestrian intention. Based on our findings, we will discuss the open problems and propose future research directions.
Figures - uploaded by Amir Rasouli
Author content
All figure content in this area was uploaded by Amir Rasouli
Content may be subject to copyright.
ResearchGate has not been able to resolve any citations for this publication.
Drivers use nonverbal cues such as vehicle speed, eye gaze, and hand gestures to communicate awareness and intent to pedestrians. Conversely, in autonomous vehicles, drivers can be distracted or absent, leaving pedestrians to infer awareness and intent from the vehicle alone. In this paper, we investigate the usefulness of interfaces (beyond vehicle movement) that explicitly communicate awareness and intent of autonomous vehicles to pedestrians, focusing on crosswalk scenarios. We conducted a preliminary study to gain insight on designing interfaces that communicate autonomous vehicle awareness and intent to pedestrians. Based on study outcomes, we developed four prototype interfaces and deployed them in studies involving a Segway and a car. We found interfaces communicating vehicle awareness and intent: (1) can help pedestrians attempting to cross; (2) are not limited to the vehicle and can exist in the environment; and (3) should use a combination of modalities such as visual, auditory, and physical.
Autonomous vehicles will have to coordinate their behavior with human road users such as drivers and pedestrians. The majority of recently proposed solutions for autonomous vehicle-to-human communication consist of introducing additional visual cues (such as lights, text and pictograms) on either the car’s exterior or as projections on the road. We argue that potential shortcomings in the visibility (due to light conditions, placement on the vehicle) and immediate understandability (learned, directive) of many of these cues make them alone insufficient in mediating multi-party interactions in the busy intersections of day-to-day traffic. Our observations of real-world human road user behavior in urban intersections indicate that movement in context is a central method of communication for coordination among drivers and pedestrians. The observed movement patterns gain meaning when seen within the context of road geometry, current road activity, and culture. While all movement communicates the intention of the driver, we highlight the use of movement as gesture, done for the specific purpose of communicating to other road users and give examples of how these influence traffic interactions. An awareness and understanding of the effect and importance of movement gestures in day-to-day traffic interactions is needed for developers of autonomous vehicles to design forms of human-vehicle communication that are effective and scalable in multi-party interactions.
In this paper, we present the preliminary results of a study that aims to investigate the role of an approaching vehicle's behavior and outer appearance in determining pedestrians' decisions while crossing a street. Concerning appearance, some vehicles are designed to look more assertive than others, and it is believed that vehicle appearance may reflect the driver's social behavior in traffic. In the case of autonomous vehicles, since the human driver no longer controls the vehicle's action, the question arises whether pedestrians treat autonomous and manually-driven vehicles differently when deciding to cross the street. We devised an experiment to determine the impact of the behavioral and physical attributes of a vehicle on pedestrians' road-crossing decisions, both for manually-driven and autonomous vehicles. Preliminary results show that in both cases, distance and speed play a dominant role in pedestrians' decision to cross a road when compared to the vehicle's size and appearance.
This paper presents a study that aimed to identify the importance of eye contact and gestures between pedestrians and drivers. A video-based observation and coding was undertaken to categorize the road-crossing and communication behavior of pedestrians and drivers in busy traffic situations where efficient negotiation is necessary. The evidence in the study suggests that eye contact does not play a major role in manual driving, that explicit communication is rare to non-existent, and that motion patterns and behaviors of vehicles play a more significant role for pedestrians in efficient traffic negotiations.
Fully automated self-driving cars, with expected benefits including improved road safety, are closer to becoming a reality. Thus, attention has turned to gauging public perceptions of these autonomous vehicles. To date, surveys have focused on the public as potential passengers of autonomous cars, overlooking other road users who would interact with them. Comparisons with perceptions of other existing vehicles are also lacking. This study surveyed almost 1000 participants on their perceptions, particularly with regards to safety and acceptance of autonomous vehicles. Overall, results revealed that autonomous cars were perceived as a “somewhat low risk“ form of transport and, while concerns existed, there was little opposition to the prospect of their use on public roads. However, compared to human-operated cars, autonomous cars were perceived differently depending on the road user perspective: more risky when a passenger yet less risky when a pedestrian. Autonomous cars were also perceived as more risky than existing autonomous trains. Gender, age and risk-taking had varied relationships with the perceived risk of different vehicle types and general attitudes towards autonomous cars. For instance, males and younger adults displayed greater acceptance. Whilst their adoption of this autonomous technology would seem societally beneficial – due to these groups’ greater propensity for taking road user risks, behaviours linked with poorer road safety – other results suggested it might be premature to draw conclusions on risk-taking and user acceptance. Future studies should therefore continue to investigate people’s perceptions from multiple perspectives, taking into account various road user viewpoints and individual characteristics.
The intelligent vehicle community has devoted considerable efforts to model driver behavior, and in particular to detect and overcome driver distraction in an effort to reduce accidents caused by driver negligence. However, as the domain increasingly shifts towards autonomous and semi-autonomous solutions, the driver is no longer integral to the decision making process, indicating a need to refocus efforts elsewhere. To this end, we propose to study pedestrian distraction instead. In particular, we focus on detecting pedestrians who are engaged in secondary activities involving their cellphones and similar handheld multimedia devices from a purely vision-based standpoint. To achieve this objective, we propose a pipeline incorporating articulated human pose estimation, followed by a soft object label transfer from an ensemble of exemplar SVMs trained on the nearest neighbors in pose feature space. We additionally incorporate head gaze features and prior pose information to carry out cellphone related pedestrian activity recognition. Finally, we offer a method to reliably track the articulated pose of a pedestrian through a sequence of images using a particle filter with a Gaussian Process Dynamical Model (GPDM), which can then be used to estimate sequentially varying activity scores at a very low computational cost. The entire framework is fast (especially for sequential data) and accurate, and easily extensible to include other secondary activities and sources of distraction.
Designing autonomous vehicles for urban environments remains an unresolved problem. One major dilemma faced by autonomous cars is understanding the intention of other road users and communicating with them. To investigate one aspect of this, specifically pedestrian crossing behavior, we have collected a large dataset of pedestrian samples at crosswalks under various conditions (e.g., weather) and in different types of roads. Using the data, we analyzed pedestrian behavior from two different perspectives: the way they communicate with drivers prior to crossing and the factors that influence their behavior. Our study shows that changes in head orientation in the form of looking or glancing at the traffic is a strong indicator of crossing intention. We also found that context in the form of the properties of a crosswalk (e.g., its width), traffic dynamics (e.g., speed of the vehicles) as well as pedestrian demographics can alter pedestrian behavior after the initial intention of crossing has been displayed. Our findings suggest that the contextual elements can be interrelated, meaning that the presence of one factor may increase/decrease the influence of other factors. Overall, our work formulates the problem of pedestrian-driver interaction and sheds light on its complexity in typical traffic scenarios.
Self-driving technologies have been increasingly developed and tested in recent years (e.g., Volvo's and Google's self-driving cars). However, only a limited number of investigations have so far been conducted into communication between self-driving cars and pedestrians. For example, when a pedestrian is about to cross a street, that pedestrian needs to know the intension of the approaching self-driving car. In the present study, we designed a novel interface known as "Eyes on a Car" to address this problem. We added eyes onto a car so as to establish eye contact communication between that car and pedestrians. The car looks at the pedestrian in order to indicate its intention to stop. This novel interface design was evaluated via a virtual reality (VR) simulated environment featuring a street-crossing scenario. The evaluation results show that pedestrians can make the correct street-crossing decision more quickly if the approaching car has the novel interface "eyes" than in the case of normal cars. In addition, the results show that pedestrians feel safer with regard to crossing a street if the approaching car has eyes and if the eyes look at them.