Pride and Prejudice: Toward an Information-Theoretic Framework for Mutually Communicative Driver Behavior Modeling
2026-06-15 • Robotics
Robotics
AI summaryⓘ
The authors study how self-driving cars and human drivers can better understand each other's intentions during lane changes to improve safety and efficiency. They propose a model where the self-driving car both signals its own intent and tries to understand the human driver's preferences, even when uncertain. Their approach uses game theory and information-sharing ideas, tested on real driving data, and shows better prediction of lane changes than methods without communication. They also confirm that listening and asking are more important than just expressing intent, and driver feedback matches their model's communication patterns.
mixed autonomy drivinglane changesBayesian persuasion gameepistemic uncertaintyinverse reinforcement learningNGSIM datasetmutual communicationdriver modelinginformation-theoretic rewardscommunication affordances
Authors
Tingjun Li, Nan Xu, Shuo Feng, Hassan Askari, Bruno Henrique Groenner Barbosa, Konghui Guo
Abstract
Mixed autonomy driving becomes unsafe and inefficient when autonomous vehicles (AVs) and human-driven vehicles (HVs) misread each other's intentions. We study this problem as implicit mutual communication in lane changes. The proposed framework models how the ego vehicle both expresses its intent and probes the other driver's preference under epistemic uncertainty. It combines a level-k Bayesian persuasion game with virtual features for proactive signaling, information-theoretic rewards for mutual communication, and adaptive weights of communication affordances. We further introduce the Pride-Inquiry (P-I) and Pride-Prejudice (P-P) planes to analyze communication intensity and tendency. The model is calibrated with a Communication-Based Multi-Agent Inverse Reinforcement Learning algorithm (C-MIRL) on the naturalistic NGSIM dataset. Compared with the non-communicative baseline, the proposed model reduces the prediction error of mandatory lane changes by up to 20% while maintaining strong generalization. Driver-In-the-Loop questionnaire scores are positively correlated with the calibrated communication variables, supporting the subjective validity of the model. The learned rewards further show that inquiry and listening affordances contribute more than pride and expression alone, and that inquiry preference varies more strongly across drivers. These results support explicit modeling of mutual communication and epistemic uncertainty in interactive driving.