Game Theory-Based Framework for Modeling Human–Vehicle Interactions on the Road

Abstract

New application domains have faded the barriers between humans and robots, introducing a new set of complexities to robotic systems. The major impediment is the uncertainties associated with human decision making, which makes it challenging to predict human behavior. A realistic model of human behavior is thus vital to capture humans’ interactive behavior with their surroundings and provide robots with reliable estimates on what is most likely to happen. Focusing on operations of connected and automated vehicles (CAVs) in areas with a high presence of human actors (i.e., pedestrians), this study creates an interactive decision-making framework to predict pedestrians’ trajectories when walking in a shared environment with vehicles and other pedestrians. It develops a game theoretical structure to approximate the movement and directional components of pedestrian motion using the theory of Nash equilibria in non-cooperative games. It also introduces a novel payoff structure to address the inherent uncertainties in human behavior. Ground truth pedestrian trajectories are then used to calibrate the game parameters and evaluate the model’s performance in approximating the motion decisions of human agents in interaction with interfering vehicles and pedestrians. The main contribution of the study is to develop an interactive human–vehicle decision-making framework toward realizing human–vehicle coexistence by capturing the effect of pedestrian–vehicle and pedestrian–pedestrian interactions on choice of walking strategies. The derived knowledge could be used in CAV navigation algorithms to provide the vehicle with more accurate predictions of pedestrian behavior, and in turn, improve CAV motion planning in human-populated areas.

The Effects of Dynamic Complexity on Drivers’ Secondary Task Scanning Behavior under a Car-Following Scenario

“…Substituting Equations ( 16)- (18) into Equation (14), the relationship between the number of icons, vehicle speed, and vehicle spacing can be expressed by Equation (19) Excessive icons increase the risk of traffic accidents during driving. To ensure driving safety, the number of icons should not exceed 24.…”

“…In this trend, the relationship between scanning times and vehicle speed can be expressed by an exponential regression model (R 2 = 0.979), the relationship between scanning times and vehicle spacing can be expressed by a negative logarithmic regression model (R 2 = 0.974), and the relationship between scanning times and the number of icons can be expressed by a positive linear regression model (R 2 = 0.783). The relationship between vehicle speed, vehicle spacing, the number of icons, and the scanning times meeting the upper limit of 95% confidence interval can be expressed by the multivariate nonlinear fitting model shown in Equation (18).…”

“…The second approach is the adaptive human-computer interaction system based on spatial three-dimensional interaction. This system uses a cooperative intelligent transportation system to collect outside information from various sensors and uses the onboard augmented reality (AR) of the windshield display screen to present the dynamic traffic information, realized through a visual display of 360 • 3D virtual space around the vehicle [13][14][15][16][17][18]. As regards the design of adaptive human-computer interaction systems, researchers believe that drivers' cognitive state and the form in which information is presented will influence the interaction effect.…”

Improved Bidirectional RRT ∗ Path Planning Method for Smart Vehicle

“…In order to adapt to dynamic traffic demands, they proposed a complex path planning strategy, which could effectively improve traffic efficiency, but did not consider the influence of information delay caused by communication uncertainty [24]. Rahmati et al considered networked autonomous vehicles and driving scenarios with high human participation and established a human-vehicle interactive decision-making framework to achieve humanvehicle coexistence, but did not combine behavior prediction and intelligent body motion planning to develop planning algorithms [25].…”

Toward Human-Centered Design of Automated Vehicles: A Naturalistic Brake Policy

“…It is expected that by 2030, nearly 50% of vehicles will still be operated by human drivers. It is, therefore, of paramount importance to equip CAVs with the required technologies that enable them to safely and efficiently operate in mixed traffic environments with both human actors and automated vehicles (Rahmati et al, 2020). CAVs in such environments need to understand humans' driving behavior and also act in a way that is safe and yet expected by surrounding human drivers.…”