Latent Dynamics-Aware OOD Monitoring for Trajectory Prediction with Provable Guarantees

TL;DR

This paper introduces a novel OOD monitoring method for trajectory prediction in safety-critical systems, using a latent dynamics-aware approach with provable guarantees, improving detection delay and robustness in real-world scenarios.

Contribution

It extends the QCD framework with a latent dynamics model and MMD-based detection, providing formal guarantees without needing explicit post-change distribution knowledge.

Findings

Reduced detection delay in real-world datasets

Robustness to heavy-tailed errors

No requirement for explicit post-change distribution

Abstract

In safety-critical Cyber-Physical Systems (CPS), accurate trajectory prediction provides vital guidance for downstream planning and control, yet although deep learning models achieve high-fidelity forecasts on validation data, their reliability degrades under out-of-distribution (OOD) scenarios caused by environmental uncertainty or rare traffic behaviors in real-world deployment; detecting such OOD events is challenging due to evolving traffic conditions and changing interaction patterns, while safety-critical applications demand formal guarantees on detection delay and false-alarm rates, motivating us-following recent work [1]-to formulate OOD monitoring for trajectory prediction as a quickest changepoint detection (QCD) problem that offers a principled statistical framework with established theory; we further observe that the real-world evolution of prediction errors under in-distribution (ID) conditions can be effectively modeled by a Hidden Markov Model (HMM), and by leveraging this structure we extend the cumulative Maximum Mean Discrepancy approach to enable detection without requiring explicit knowledge of the post-change distribution while still admitting provable guarantees on delay and false alarms, with experiments on three real-world driving datasets demonstrating reduced detection delay and robustness to heavy-tailed errors and unknown post-change conditions.