MEGA Hub

G-MARK: Grounded Multi-Agent Reasoning for Cooperative Driving via Knowledge Graphs

Authors

Do you know Bhavya Gupta?You can claim authorship or link another user.Do you know Onat Gungor?You can claim authorship or link another user.Do you know Tajana Rosing?You can claim authorship or link another user.

Abstract

Autonomous driving systems must operate under partial observability, where safety-critical objects may be occluded or visible only to neighboring connected vehicles. Vehicle-to-vehicle cooperation can reduce this uncertainty, but existing cooperative driving methods often compress multi-agent evidence into latent features or hidden multimodal states. As a result, they obscure which agent observed each object, whether the object is visible to the ego vehicle, and how conflicting evidence affects downstream decisions. We propose G-MARK, a grounded multi-agent reasoning framework that converts cooperative object-centric observations into explicit provenance-aware knowledge graphs (KGs). The resulting KGs preserve object hypotheses together with their source attribution, ego-versus-partner visibility, uncertainty, conflicts, spatial relations, and planning-relevant context. G-MARK then derives a shared feature representation from these KGs, enabling lightweight task heads to support object reasoning, motion prediction, control selection, and trajectory forecasting. Compared with the state-of-the-art baseline, GMARK improves occlusion reasoning accuracy by 42.2%, reduces control-selection error by 13.1%, and achieves comparable trajectory-planning accuracy with a 25.6x smaller structured communication payload. Our code is available at https://github.com/bhavyagupta98/g-mark.

Community

00

Publication notes

Author note
Accepted for oral presentation at the 25th IEEE International Conference on Machine Learning and Applications (ICMLA'26)