Conformal Prediction for Robotics

Overview

This research direction explores the application of conformal prediction techniques in robotics. Conformal prediction provides a framework for quantifying uncertainty in predictions, which is crucial for robotic systems operating in dynamic and uncertain environments. While these techniques are widely studied in machine learning, their application to robotics is still emerging. This work aims to bridge that gap by developing conformal prediction methods tailored for robotic applications such as perception, control, and decision-making.

Data-Driven Non-Conformity Scores

One critical problem in conformal prediction is the design of non-conformity scores, which measure how well a new data point conforms to the training data. Traditional approaches often rely on simple distance metrics or residuals, which may not capture the complex relationships in robotic data. This research direction focuses on developing data-driven non-conformity scores that leverage machine learning models to better capture the underlying structure of the data, leading to more accurate and informative prediction sets.

A python library for data-driven non-conformity scores is available at conformal_region_designer.

Selected publications:

Physics Constrained Motion Prediction with Uncertainty Quantification [1]
Renukanandan Tumu, Lars Lindemann, Truong X. Nghiem, Rahul Mangharam
Intelligent Vehicles (IV), 2023
Multi-Modal Conformal Prediction Regions by Optimizing Convex Shape Templates [2]
Renukanandan Tumu, Matthew Cleaveland, Rahul Mangharam, George Pappas, Lars Lindemann
Learning for Dynamics and Control (L4DC), 2024
AdaptNC: Adaptive Nonconformity Scores for Uncertainty-Aware Autonomous Systems in Dynamic Environments [3]
Renukanandan Tumu, Aditya Singh, Rahul Mangharam
Under Review, International Conference on Machine Learning (ICML), 2026

Conformal Prediction in Multi-Agent Settings

Another critical problem in the application of conformal prediction to robotics is overcoming distribution shifts that arise in multi-agent settings. In such scenarios, the behavior of one agent can significantly influence the environment and the data distribution observed by other agents. This research direction focuses on developing conformal prediction methods that adapt to these distribution shifts, ensuring that prediction sets remain valid and informative even as interactions evolve.

Selected publications:

Conformal Off-Policy Prediction for Multi-Agent Systems [4]
Tom Kuipers, Renukanandan Tumu, Shuo Yang, Milad Kazemi, Rahul Mangharam, Nicola Paoletti
Conference on Decision and Control (CDC), 2024
AdaptNC: Adaptive Nonconformity Scores for Uncertainty-Aware Autonomous Systems in Dynamic Environments [3]
Renukanandan Tumu, Aditya Singh, Rahul Mangharam
Under Review, ICML, 2026

References

2026

AdaptNC: Adaptive Nonconformity Scores for Uncertainty-Aware Autonomous Systems in Dynamic Environments

Renukanandan Tumu, Aditya Singh, and Rahul Mangharam

Feb 2026

Abs PDF

Rigorous uncertainty quantification is essential for the safe deployment of autonomous systems in unconstrained environments. Conformal Prediction (CP) provides a distribution-free framework for this task, yet its standard formulations rely on exchangeability assumptions that are violated by the distribution shifts inherent in real-world robotics. Existing online CP methods maintain target coverage by adaptively scaling the conformal threshold, but typically employ a static nonconformity score function. We show that this fixed geometry leads to highly conservative, volume-inefficient prediction regions when environments undergo structural shifts. To address this, we propose \textbfAdaptNC, a framework for the joint online adaptation of both the nonconformity score parameters and the conformal threshold. AdaptNC leverages an adaptive reweighting scheme to optimize score functions, and introduces a replay buffer mechanism to mitigate the coverage instability that occurs during score transitions. We evaluate AdaptNC on diverse robotic benchmarks involving multi-agent policy changes, environmental changes and sensor degradation. Our results demonstrate that AdaptNC significantly reduces prediction region volume compared to state-of-the-art threshold-only baselines while maintaining target coverage levels.

2024

Multi-modal conformal prediction regions by optimizing convex shape templates

Renukanandan Tumu^*, Matthew Cleaveland^*, Rahul Mangharam, and 2 more authors

In Proceedings of the 6th Annual Learning for Dynamics & Control Conference, Jul 2024

Abs PDF

Conformal prediction is a statistical tool for producing prediction regions for machine learning models that are valid with high probability. A key component of conformal prediction algorithms is a non-conformity score function that quantifies how different a model’s prediction is from the unknown ground truth value. Essentially, these functions determine the shape and the size of the conformal prediction regions. However, little work has gone into finding non-conformity score functions that produce prediction regions that are multi-modal and practical, i.e., that can efficiently be used in engineering applications. We propose a method that optimizes parameterized shape template functions over calibration data, which results in non-conformity score functions that produce prediction regions with minimum volume. Our approach results in prediction regions that are multi-modal, so they can properly capture residuals of distributions that have multiple modes, and practical, so each region is convex and can be easily incorporated into downstream tasks, such as a motion planner using conformal prediction regions. Our method applies to general supervised learning tasks, while we illustrate its use in time-series prediction. We provide a toolbox and present illustrative case studies of F16 fighter jets and autonomous vehicles, showing an up to 68% reduction in prediction region area.
Conformal Off-Policy Prediction for Multi-Agent Systems

Tom Kuipers, Renukanandan Tumu, Shuo Yang, and 3 more authors

In 2024 IEEE 63rd Conference on Decision and Control (CDC), 2024

DOI

2023

Physics Constrained Motion Prediction with Uncertainty Quantification

Renukanandan Tumu, Lars Lindemann, Truong Nghiem, and 1 more author

In 2023 IEEE Intelligent Vehicles Symposium (IV), Jun 2023

Abs DOI HTML

Predicting the motion of dynamic agents is a critical task for guaranteeing the safety of autonomous systems. A particular challenge is that motion prediction algorithms should obey dynamics constraints and quantify prediction uncertainty as a measure of confidence. We present a physics-constrained approach for motion prediction which uses a surrogate dynamical model to ensure that predicted trajectories are dynamically feasible. We propose a two-step integration consisting of intent and trajectory prediction subject to dynamics constraints. We also construct prediction regions that quantify uncertainty and are tailored for autonomous driving by using conformal prediction, a popular statistical tool. Physics Constrained Motion Prediction achieves a 41% better ADE, 56% better FDE, and 19% better IoU over a baseline in experiments using an autonomous racing dataset.