Prediction uncertainty-aware planning using deep ensembles and trajectory optimisation

2025-02-14

AI Summary

Deterministic trajectory prediction in crowded pedestrian scenes often leads to overconfident motion planning, increasing collision risk. Method: This paper proposes a probabilistic-aware end-to-end motion planning framework. Its core innovation is the first explicit incorporation of prediction uncertainty—estimated via a deep ensemble model—into nonlinear trajectory optimization as dynamic safety constraints, enabling uncertainty-driven safe planning. The approach jointly integrates probabilistic trajectory forecasting, uncertainty quantification, and constraint-embedded optimization. Results: Evaluated on real-world pedestrian datasets, the method significantly improves constraint satisfaction rates. It successfully achieves offline safe navigation in challenging scenarios—including narrow corridors and out-of-distribution pedestrian trajectories—demonstrating strong robustness and generalization capability.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Planning, Routing, and Scheduling: Planning under Uncertainty
• Intelligent Robots: Motion and Path Planning

Application Category

• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Human motion is stochastic and ensuring safe robot navigation in a pedestrian-rich environment requires proactive decision-making. Past research relied on incorporating deterministic future states of surrounding pedestrians which can be overconfident leading to unsafe robot behaviour. The current paper proposes a predictive uncertainty-aware planner that integrates neural network based probabilistic trajectory prediction into planning. Our method uses a deep ensemble based network for probabilistic forecasting of surrounding humans and integrates the predictive uncertainty as constraints into the planner. We compare numerous constraint satisfaction methods on the planner and evaluated its performance on real world pedestrian datasets. Further, offline robot navigation was carried out on out-of-distribution pedestrian trajectories inside a narrow corridor

Problem

• Predicting stochastic human motion accurately
• Ensuring safe robot navigation in crowds
• Integrating predictive uncertainty into planning

Innovation

• Deep ensemble for uncertainty prediction
• Probabilistic trajectory optimization
• Real-world pedestrian dataset testing