Agentic Driving Coach: Robustness and Determinism of Agentic AI-Powered Human-in-the-Loop Cyber-Physical Systems

TL;DR

This paper introduces a reactor-model-of-computation approach using Lingua Franca to improve robustness and determinism in agentic AI-powered human-in-the-loop cyber-physical systems, demonstrated through a driving coach case study.

Contribution

It presents a novel reactor-model-of-computation framework implemented with Lingua Franca to enhance determinism in agentic AI HITL CPS, addressing a key challenge in the field.

Findings

LF-based system reveals practical challenges in reintroducing determinism.

Case study demonstrates the approach in an agentic driving coach.

Pathways to improve robustness and control in agentic HITL CPS are identified.

Abstract

Foundation models, including large language models (LLMs), are increasingly used for human-in-the-loop (HITL) cyber-physical systems (CPS) because foundation model-based AI agents can potentially interact with both the physical environments and human users. However, the unpredictable behavior of human users and AI agents, in addition to the dynamically changing physical environments, leads to uncontrollable nondeterminism. To address this urgent challenge of enabling agentic AI-powered HITL CPS, we propose a reactor-model-of-computation (MoC)-based approach, realized by the open-source Lingua Franca (LF) framework. We also carry out a concrete case study using the agentic driving coach as an application of HITL CPS. By evaluating the LF-based agentic HITL CPS, we identify practical challenges in reintroducing determinism into such agentic HITL CPS and present pathways to address them.