Sample-Efficient Safety Assurances using Conformal Prediction
This work addresses safety assurance for robotics deployments, providing a sample-efficient method to detect unsafe situations with provable guarantees, which is incremental but practical for real-world applications.
The paper tackles the problem of ensuring safety in high-stakes robotics by developing a framework that combines conformal prediction with simulators to tune warning systems, achieving a provable false negative rate of ε using as few as 1/ε data points, as demonstrated in driver warning and robotic grasping applications with low false positive rates.
When deploying machine learning models in high-stakes robotics applications, the ability to detect unsafe situations is crucial. Early warning systems can provide alerts when an unsafe situation is imminent (in the absence of corrective action). To reliably improve safety, these warning systems should have a provable false negative rate; i.e. of the situations that are unsafe, fewer than $ε$ will occur without an alert. In this work, we present a framework that combines a statistical inference technique known as conformal prediction with a simulator of robot/environment dynamics, in order to tune warning systems to provably achieve an $ε$ false negative rate using as few as $1/ε$ data points. We apply our framework to a driver warning system and a robotic grasping application, and empirically demonstrate guaranteed false negative rate while also observing low false detection (positive) rate.