EB-MBD: Emerging-Barrier Model-Based Diffusion for Safe Trajectory Optimization in Highly Constrained Environments

2025-10-08

AI Summary

In highly constrained environments, model-based diffusion methods suffer from sharply degraded safe trajectory optimization performance due to low sampling efficiency in Monte Carlo score estimation. To address this, we propose a progressive barrier function embedding mechanism. Our approach innovatively integrates an interior-point–inspired dynamic barrier function into the diffusion process and adaptively schedules barrier parameters via sampling activity analysis—thereby preventing sample starvation while eliminating costly projection operations. Crucially, constraints are introduced progressively and smoothly without increasing model complexity. Experiments on 2D obstacle avoidance and 3D underwater robotic arm tasks demonstrate that our method significantly reduces trajectory cost and improves computational efficiency by several orders of magnitude over conventional projection-based approaches. This work establishes an efficient, scalable new paradigm for safety-critical planning under stringent constraints.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Constraint Satisfaction and Optimization: Constraint Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Security and Privacy: Large-scale security measurements

Abstract

We propose enforcing constraints on Model-Based Diffusion by introducing emerging barrier functions inspired by interior point methods. We show that constraints on Model-Based Diffusion can lead to catastrophic performance degradation, even on simple 2D systems due to sample inefficiency in the Monte Carlo approximation of the score function. We introduce Emerging-Barrier Model-Based Diffusion (EB-MBD) which uses progressively introduced barrier constraints to avoid these problems, significantly improving solution quality, without the need for computationally expensive operations such as projections. We analyze the sampling liveliness of samples each iteration to inform barrier parameter scheduling choice. We demonstrate results for 2D collision avoidance and a 3D underwater manipulator system and show that our method achieves lower cost solutions than Model-Based Diffusion, and requires orders of magnitude less computation time than projection based methods.

Problem

• Improving trajectory optimization in constrained environments
• Overcoming sample inefficiency in diffusion-based methods
• Reducing computational cost without projection operations

Innovation

• Emerging barrier functions enforce constraints on diffusion
• Progressively introduced barriers avoid catastrophic performance degradation
• Barrier parameter scheduling informed by sampling liveliness analysis