The Constitutional Controller: Doubt-Calibrated Steering of Compliant Agents

TL;DR

This paper introduces the Constitutional Controller (CoCo), a neuro-symbolic framework that improves autonomous agents' safety and compliance in uncertain environments by integrating probabilistic logic with deep learning and employing self-doubt measures.

Contribution

The paper presents CoCo, a novel neuro-symbolic system that combines probabilistic logic with deep learning to enhance safety and rule compliance in autonomous agents.

Findings

CoCo effectively reasons over constraints in shared traffic spaces.

Self-doubt helps agents assess their confidence in decision-making.

Real-world aerial mobility experiments demonstrate improved safety and compliance.

Abstract

Ensuring reliable and rule-compliant behavior of autonomous agents in uncertain environments remains a fundamental challenge in modern robotics. Our work shows how neuro-symbolic systems, which integrate probabilistic, symbolic white-box reasoning models with deep learning methods, offer a powerful solution to this challenge. This enables the simultaneous consideration of explicit rules and neural models trained on noisy data, combining the strength of structured reasoning with flexible representations. To this end, we introduce the Constitutional Controller (CoCo), a novel framework designed to enhance the safety and reliability of agents by reasoning over deep probabilistic logic programs representing constraints such as those found in shared traffic spaces. Furthermore, we propose the concept of self-doubt, implemented as a probability density conditioned on doubt features such as travel velocity, employed sensors, or health factors. In a real-world aerial mobility study, we demonstrate CoCo's advantages for intelligent autonomous systems to learn appropriate doubts and navigate complex and uncertain environments safely and compliantly.