Ali Baheri
Assistant Professor | Rochester Institute of Technology
I am an assistant professor in the Department of Mechanical Engineering at Rochester Institute of Technology where I lead the Safe AI Lab. Our lab focuses on research at the intersection of autonomy, controls, and machine learning. Before joining RIT, I was a visiting scholar at Stanford Intelligent Systems Laboratory (SISL). Prior to that, I was an assistant professor (in the research track) at West Virginia University. Before that, I was a postdoctoral research fellow at the University of Michigan—Ann Arbor. I received my Ph.D. at the University of North Carolina at Charlotte in 2018.
Our vision is to make progress toward safe, certified, and efficient autonomy with the application of intelligent systems on the ground or in the air. To achieve that goal, we leverage tools from artificial intelligence, data-driven optimization techniques, and control theory. Our work has been generously funded by NSF, FAA, and NASA.
news
Our paper titled “Wasserstein Stability of Contracting Flows: Effective Rates, Euler Self-Correction, and Noise Tightening” has been accepted for the 2026 IEEE Conference on Decision and Control (CDC)!
Our paper titled “How Much Do Your Models Disagree? Adaptive MPC Safety from Ensemble Uncertainty” has been accepted for publication in IEEE Open Journal of Control Systems!
Our paper titled Flow-Corrected Thompson Sampling for Non-Stationary Contextual Bandits has been accepted to the Continual RL Workshop at RLC 2026 and selected for an oral presentation!
Our paper titled Blending Optimism and Pessimism via Wasserstein Barycenters for Continuous Control has been published in IEEE Control Systems Letters!
Our paper titled Logic-Guided Vector Fields for Constrained Generative Modeling has been accepted at NeuS 2026 and selected for a spotlight talk!
Our paper titled Efficient Counterexample Generation for Control Systems Using Multi-Fidelity Bayesian Optimization has been published in IEEE Access!
Our paper titled “Metriplectic Conditional Flow Matching for Structure-Preserving Dynamics Learning” has been accepted to the 2026 European Control Conference (ECC)!
Three papers accepted to the American Control Conference (ACC) 2026!
Our paper titled Geometry-Aware Backdoor Attacks: Leveraging Curvature in Hyperbolic Embeddings has been accepted by the NeurIPS 2025 Workshop on Non-Euclidean Foundation Models and Geometric Learning!
Our paper titled Metriplectic Conditional Flow Matching for Dissipative Dynamics has been accepted by the NeurIPS 2025 Workshop on Dynamics at the Frontiers of Optimization, Sampling, and Games (DynaFront)!
Our paper titled Hierarchical Neuro-Symbolic Decision Transformer has been accepted by 19th Conference on Neurosymbolic Learning and Reasoning (NeSy 2025)!
Our paper titled Wasserstein-Barycenter Consensus for Cooperative Multi-Agent Reinforcement Learning has been accepted by the ICML 2025 Workshop on Multi-Agent Systems in the Era of Foundation Models!
Our paper titled Implicit Constraint‑Aware Off‑Policy Correction for Offline Reinforcement Learning has been accepted by the RSS 2025 Workshop on Out-of-Distribution Generalization in Robotics!
Our paper titled Multi-Fidelity Temporal Reasoning: A Stratified Logic for Cross-Scale System Specifications has been accepted by Logics!
Our paper titled Conformal prediction across scales: Finite-sample coverage with hierarchical efficiency has been accepted by Results in Applied Mathematics!
Our paper titled Distributionally Robust Lyapunov-Barrier Networks for Safe and Stable Control Under Uncertainty has been accepted by the Results in Control and Optimization (RICO)!
Our paper titled “WAVE: Wasserstein Adaptive Value Estimation for Actor-Critic Reinforcement Learning” has been accepted by the 2025 Learning for Dynamics & Control Conference (L4DC)!
Our paper titled Multilevel Constrained Bandits: A Hierarchical Upper Confidence Bound Approach with Safety Guarantees has been accepted by the Mathematics!
Our paper titled Forward Variable Selection Enables Fast and Accurate Dynamic System Identification with Karhunen-Loève Decomposed Gaussian Processes has been accepted by the PLOS One! (Joint work with our colleagues at West Virginia University)
Our paper titled BEACON: A Bayesian Evolutionary Approach for Counterexample Generation of Control Systems has been accepted by IEEE Access!
Our paper titled A Survey on Reinforcement Learning in Aviation Applications has been accepted by the Engineering Applications of Artificial Intelligence journal! (Joint work with our colleagues at George Washington University)
Our paper titled “Optimal Transport-Assisted Risk-Sensitive Q-Learning” has been accepted by the Towards Safe Autonomy workshop at RSS 2024!
Our paper titled Concurrent Learning of Control Policy and Unknown Safety Specifications in Reinforcement Learning has been accepted by the IEEE Open Journal of Control Systems!
New arXiv paper out: The Synergy Between Optimal Transport Theory and Multi-Agent Reinforcement Learning
Received a seed grant from RIT AI Seed Fund!
Our paper titled “Exploring the Role of Simulator Fidelity in the Safety Validation of Learning-Enabled Autonomous Systems” has been accepted by the AI Magazine!
Our paper titled “LLMs-Augmented Contextual Bandit” has been accepted by the Foundation Models for Decision Making workshop at NeurIPS 2023!
Our paper titled “Understanding Reward Ambiguity Through Optimal Transport Theory in Inverse Reinforcement Learning” has been accepted by the Optimal Transport and Machine Learning workshop at NeurIPS 2023!
Our paper titled “Risk-Aware Reinforcement Learning Through Optimal Transport Theory” has been accepted by the 3rd RL-CONFORM workshop at IROS 2023!
Both papers submitted by the Safe AI Lab to the “Bridging the Gap Between AI Planning and Reinforcement Learning (PRL)” workshop at ICAPS 2023 have been accepted!
Our paper titled “Falsification of Learning-Based Controllers through Multi-Fidelity Bayesian Optimization” has been accepted for the 2023 European Control Conference (ECC)!
I gave a talk at the AAAI 2023 New Faculty Highlights Program entitled “On the Role of Fidelity in the Safety Evaluation of Learning-Based Autonomous Systems”.
Selected for the AAAI 2023 New Faculty Highlights Program.
We co-organize “Machine Learning for Autonomous Driving (ML4AD)” workshop at NeurIPS, 2022.
We co-organize the “NASA EPSCOR HACKWEEK” workshop at West Virginia University.
I gave a talk at the ICRA 2022 Workshop on the Verification of Autonomous Systems (VAS).
I gave a talk at the University of North Texas entitled “Safe Decision-Making in Evolving Environments for Safety-Critical Autonomous Systems”.
Our project “Safety Validation of Autonomous Systems from Multiple Sources of Information” has been funded by the National Science Foundation (NSF).
Our recent work on safe reinforcement learning with mixture density network has been accepted by the Results in Control and Optimization journal.Â
Our project “Safety Verification Framework for Learning-based Aviation Systems” has been funded by the Federal Aviation Administration (FAA).
Our project “Verification of Multi-Agent Autonomous Planning and Control” has been funded by the West Virginia University Research Office Program.
We co-organize “Machine Learning for Autonomous Driving (ML4AD)” workshop at NeurIPS, 2021.
Our project “Fault Diagnosis for Safety-Critical Autonomous Systems using Reinforcement Learning” has been funded by NASA.
Our project “Black-box Verification of Autonomous Systems using Modular Reinforcement Learning” has been funded by NASA WV Space Grant Consortium.
I gave a talk at Ford Motor Company.
Our project “Robust Autonomy through Experimentally Infused Decision-Making” has been funded by NASA WV Space Grant Consortium.
One paper accepted to 2021 IEEE Intelligent Vehicles Symposium.
I gave a talk at the University of New Mexico.
Two papers are accepted for the 2020 American Control Conference (ACC).
I joined West Virginia University as an Assistant Professor of Research.