IANN-MPPI: Interaction-Aware Neural Network-Enhanced Model Predictive Path Integral Approach for Autonomous Driving

2025-07-16

AI Summary

To address conservative and failed motion planning in autonomous vehicles operating in dense traffic—caused by insufficient interactive perception—this paper proposes an interactive motion planning method. The approach tightly integrates neural-network-based interactive behavior prediction with the Model Predictive Path Integral (MPPI) control framework, enabling online modeling of surrounding vehicles’ responses and closed-loop cooperative trajectory optimization. Additionally, a spline-based prior distribution is introduced to improve sampling efficiency under lane constraints. Unlike conventional decoupled prediction-and-planning pipelines, our method explicitly models the bidirectional interaction mechanism: “ego-vehicle action → neighboring vehicle response.” Extensive experiments in dense on-ramp merging scenarios demonstrate significant improvements in merge success rate and traffic throughput, validating enhanced proactive interaction and cooperative driving capabilities.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Planning, Routing, and Scheduling: Planning with Language Models
• Intelligent Robots: Motion and Path Planning

Application Category

• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Motion planning for autonomous vehicles (AVs) in dense traffic is challenging, often leading to overly conservative behavior and unmet planning objectives. This challenge stems from the AVs' limited ability to anticipate and respond to the interactive behavior of surrounding agents. Traditional decoupled prediction and planning pipelines rely on non-interactive predictions that overlook the fact that agents often adapt their behavior in response to the AV's actions. To address this, we propose Interaction-Aware Neural Network-Enhanced Model Predictive Path Integral (IANN-MPPI) control, which enables interactive trajectory planning by predicting how surrounding agents may react to each control sequence sampled by MPPI. To improve performance in structured lane environments, we introduce a spline-based prior for the MPPI sampling distribution, enabling efficient lane-changing behavior. We evaluate IANN-MPPI in a dense traffic merging scenario, demonstrating its ability to perform efficient merging maneuvers. Our project website is available at https://sites.google.com/berkeley.edu/iann-mppi

Problem

• Enhance AV motion planning in dense traffic
• Predict interactive agent responses to AV actions
• Improve lane-changing efficiency in structured environments

Innovation

• Neural network predicts agent reactions interactively
• Spline-based prior enhances lane-changing efficiency
• MPPI integrates interactive predictions for trajectory planning