Adaptive uncertainty intervals improve safe control under uncertainty
Safety-Critical Control under Uncertainty via Adaptive Conformal Quantile Prediction Intervals
Robotics
Summary
Safety-critical systems, like self-driving cars or robots, need to handle uncertainty carefully to avoid accidents. The authors introduce a new method that creates smart, on-the-spot guesses about uncertainty that change depending on the situation, instead of making broad or fixed assumptions. Their approach uses a technique called adaptive conformal prediction combined with quantile regression to build these useful uncertainty intervals. This helps control systems stay safe while being less cautious and more effective. They test their method in simulations and provide theoretical backing.
What this means in practice
- •For autonomous vehicle engineers: Use adaptive, state-dependent uncertainty intervals to improve safety constraints in vehicle control algorithms under uncertain conditions.
- •For robotics control teams: Incorporate calibrated uncertainty predictions into robot motion controllers to reduce overly cautious behavior while maintaining high safety levels.
Authors
Hao Zhou, Yanze Zhang, Yiwei Lyu, Wenhao Luo
Abstract
Safety-critical control under uncertainty requires uncertainty representations that are both statistically valid (for certifiable performance) and compatible with enforceable safety constraints. However, existing methods often assume particular distributions of uncertainty for provable safety guarantees or establish symmetric and input-agnostic prediction intervals for robust safety, which can lead to misaligned or overly conservative safety constraints in control synthesis. In this paper, we introduce a novel safe control framework with adaptive uncertainty quantification that constructs calibrated and state-dependent prediction intervals to enable high-probability safety guarantees, while improving constrained control performance. The framework leverages adaptive conformal prediction (ACP) and extends it with conformal quantile regression (CQR) to capture distribution-free, asymmetric uncertainty intervals with certifiable probabilistic coverage, and integrates the resulting uncertainty sets into a probabilistic control barrier function formulation to enforce robust safety with reduced conservativeness. This yields uncertainty-aware safe control constraints that can be incorporated within a model predictive control(MPC) framework to provide provably safe behaviors with high probability. Simulation and theoretical results are provided to demonstrate the effectiveness of our approach.