MIND-CAVs: Multi-Intelligence Negotiation and Decision System for CAVs based on Intent-Driven Autonomy

2026-07-16

AI Summary

This work addresses the lack of high-level, intention-driven global coordination in current connected autonomous driving systems, which struggle to resolve conflicting driving intents efficiently and safely. The authors propose a multi-agent cooperative decision-making framework grounded in structured driving intentions: vehicles abstract their perceptions into structured intents and share them via V2X communication; roadside edge servers arbitrate conflicts by fusing learned models with rule-based reasoning; and a cloud platform enables decision auditing and continual learning. By introducing structured driving intent as the fundamental unit of coordination and establishing a closed-loop cloud-edge-vehicle architecture, the approach significantly reduces maneuver completion time, minimizes unsafe proximities and unnecessary braking in multi-lane highway scenarios, and outperforms baseline methods including isolated autonomy, first-come-first-served scheduling, and multi-agent reinforcement learning.

Abstract

Modern autonomous vehicles largely operate as isolated agents: they rely on on-board perception and decision modules and broadcast Basic Safety Messages (BSMs) that expose only low-level kinematic state. While existing cooperative driving frameworks enable limited sensor sharing, they rarely communicate high-level maneuver intentions, and edge computing is primarily used for content delivery rather than decision arbitration. As a result, current connected autonomy lacks a principled mechanism for making globally consistent, intent-aware coordination decisions across vehicles. To address this gap, we propose MIND-CAVs, a Multi-Intelligence Negotiation and Decision framework for connected autonomous vehicles (CAVs) based on intent-driven autonomy. Each vehicle abstracts raw sensor observations into structured intent representations, exchanges them over V2X links, and receives globally consistent coordination plans from roadside edge servers. Edge agents combine learned and rule-based arbitration mechanisms to negotiate conflicting intents among vehicles, while a cloud platform records decisions for auditing and continual retraining. We implement MIND-CAVs in a CARLA-based AI-in-the-loop platform and evaluate it in multi-lane highway scenarios involving conflicting maneuvers and route-constrained exits. Experimental results show improved maneuver completion time and reduced unsafe proximity and unnecessary braking compared with isolated autonomy, first-come-first-served arbitration, and multi-agent reinforcement learning baselines.

Problem

• connected autonomous vehicles
• intent-driven autonomy
• cooperative driving
• decision coordination
• V2X communication

Innovation

• intent-driven autonomy
• multi-intelligence negotiation
• connected autonomous vehicles
• edge-based arbitration
• V2X intent exchange