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

Deeksha Prahlad; Daniel Fan; Hokeun Kim

cs.AIApr 13, 2026

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

Authors: Deeksha Prahlad, Daniel Fan, Hokeun Kim

Organizations: Arizona State University Tempe, AZ, United States

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.

Figures & tables

Fig. 1: Challenges faced by an agentic driving coach as an agentic AI-powered HITL CPS. (a) The agent delivers an instruction to the driver too late (at time

t+5

). (b) The agent’s instruction is incorrect, although delivered timely at time

t

Fig. 2: HITL CPS depicting the interaction between the three key components, (a) agentic coach, (b) human driver, and (c) physical plant that includes the environment and car.

Fig. 3: Our proposed reactor model implemented using Lingua Franca, targeting the stop sign scenario, consists of four key reactors: (a) Coach , (b) Driver , (c) Car , and (d) Environment . Coach includes two sub-reactors: (e) LLMInference and (f) Planner .

Driver behavior	Behavior Types	Details
Accelerator	Coasting	$a=-0.1$ $\mathrm{m/s^{2}}$
	Cruise	$a=0.1$ $\mathrm{m/s^{2}}$
	Normal Acceleration	$a=2$ $\mathrm{m/s^{2}}$
	Strong Acceleration	$a=4$ $\mathrm{m/s^{2}}$
Brake	Gentle braking	$a=-3.0$ $\mathrm{m/s^{2}}$
Brake	Emergency braking	$a=-9.0$ $\mathrm{m/s^{2}}$

TABLE I: Abstracted driver behaviors used in our agentic driving coach model.

a

represents the acceleration of the car.

Notation	Description
$v_{n}$	Velocity of the car at time $n$
$v_{n+1}$	Velocity of the car at time $n+1$
$s_{n}$	Displacement of the car at time $n$
$s_{n+1}$	Displacement of the car at time $n+1$
$a$	Net acceleration of the car
$\Delta t$	Time interval between two consecutive time steps

TABLE II: Notations used in the Equations ( 2 ) and ( 3 ).

Control Signal	Description
NONE	No intervention & driver
	behavior remains within velocity bounds.
WARNING	Driver behavior deviates
	from velocity bounds.
ACTUATE	Safety limits are violated &
	the system triggers actuation.

TABLE III: Control signals generated by the LLM for mode transitions in the Planner reactor.

Fig. 4: The Planner reactor as a modal model where transitions are triggered by control signals defined in TABLE III .

Fig. 5: An example structured prompt provided for LLMInference to generate a control signal (denoted as TOKEN) and a driving instruction for the stop sign scenario.

Fig. 6: Three scenarios to evaluate the implementation of the proposed approach. (a) Stop sign, (b) Speed changing, and (c) Lane changing.

Fig. 7: Evaluation results illustrating velocity vs. displacement and time for three Llama 3 models (1B, 8B, 70B) in a stop sign scenario. Blue lines show car velocity; solid/dashed red lines show safety bounds of car velocity. Red X marks deadline violations in LLMInference . Green and orange triangles indicate coach instructions and actuation commands, respectively.

Fig. 8: Evaluation results with Llama 3 (8B vs. 70B) in speed and lane change scenarios, illustrated by velocity vs. displacement and time.

Human-in-the-Loop AI

Human-in-the-loop AI combines human feedback, oversight, and decision-making with machine learning and autonomous systems.

See all 289 papers

Neuro-Agentic Control: A Deep Learning-based LLM-Powered Agentic AI Framework for Controlling Security Controls
Jul 10, 2026Saroj Gopali, Bipin Chhetri, Deepika Giri +2Industrial Control System Security Language Model-Based Control
Department of Computer Science1, Texas Tech University1 · Cumberland University2 · Advanced Academic Programs Science3, Johns Hopkins University3
When Agents Control Robots: A Zero Trust Policy Model for Agentic Cyber-Physical Systems
May 25, 2026Tharindu Ranathunga, Kavishka Fernando, Susan ReaRobotic Control Cyber-Physical Systems
Nimbus Centre, Munster Technological University, Cork, Ireland
Harness as an Asset: Enforcing Determinism via the Convergent AI Agent Framework (CAAF)
Apr 18, 2026Tianbao ZhangSafety-Critical Control LLM Guardrails
Independent Researcher

Jul 10, 2026J/K move · Enter open · S save

Neuro-Agentic Control: A Deep Learning-based LLM-Powered Agentic AI Framework for Controlling Security Controls

Saroj Gopali, Bipin Chhetri, Deepika Giri, Sima Siami-Namini+1

Department of Computer Science1, Texas Tech University1 · Cumberland University2 · Advanced Academic Programs Science3, Johns Hopkins University3

Industrial Control System Security Language Model-Based Control Safety-Critical Control Cybersecurity AI Agent Security AI Agent Safety

Open paper AskPDF arXiv

Cyberattacks on operational technology are increasingly causing costly downtime and physical damage, exposing the limitations of traditional rule-based monitoring in industrial IoT environments. While Large Language Models (LLMs) have strong semantic reasoning abilities to assist in decision support, their hallucinatory nature presents unacceptable safety liabilities for closed-loop control. This paper introduces a neuro-agentic control framework, a novel architecture that couples an LLM-based planner (i.e., such as Gemini 2.5 Flash-Lite) with a pre-trained Time-Series Foundation Model (TimesFM), to achieve physics-grounded autonomous defense. The paper introduces a

Counterfactual Physics Injection'' mechanism that simulates the impact of LLM-proposed interventions within the numerical latent space of the foundation model before actuation, while allowing the system to reject hallucinatory or unsafe actions. Evaluated on an industrial dataset (e.g., the Secure Water Treatment (SWaT)) in the context of stochastic attack scenarios, the framework exhibited better performance compared to LSTM and TCN baselines. The Neuro-Agentic Loop prevented five breaches (33.3%) below the threshold versus LSTM (26.7%) and TCN (13.3%), with zero physically invalid (hallucinated) actions executed. These results demonstrate the efficacy of using foundation models as deterministic

Sentinels'' to safeguard agentic AI in critical infrastructure.

Abstract

Human-in-the-Loop AI

Explore similar work

Neuro-Agentic Control: A Deep Learning-based LLM-Powered Agentic AI Framework for Controlling Security Controls

When Agents Control Robots: A Zero Trust Policy Model for Agentic Cyber-Physical Systems

Harness as an Asset: Enforcing Determinism via the Convergent AI Agent Framework (CAAF)