Agentic AI · Autonomy · Industry leadership
Al Arab, Ph.D.
Applied ML/AI Leader
Executive Director, GenAuto AI · NYU Tandon and CCNY
I lead applied ML/AI for agentic autonomy: systems that plan, act, and must be scored before they ship. My organizations convert operator and customer intent into evaluation, launch, and monitoring systems for vehicles and industrial machines.
- 10+ yearsApplied AI in industry
- Agentic AIPlan, act, and evaluate
- 5U.S. patents
- IEEESenior Member
Capabilities
Agentic systems for automotive and industrial operations
-
01
Agentic AI
Design and govern AI agents that close the loop: perceive, plan, and act under operational constraints. Includes policy agents for driving and robot behavior, generative agents for scenario and risk discovery, and evaluation agents that score closed-loop performance.
-
02
Behavior validation
Translate operator and customer expectations into critics, metrics, and human baselines. Evaluate agent behavior on safety, comfort, and intent, not specification compliance alone.
-
03
Offboard ML
Train and operate vision and time-series models that infer agent and scene behavior. Feed those signals into validation pipelines, continuous release, and in-service monitoring.
-
04
Release-ready V&V
Architect simulation-led evaluation that supports launch decisions. Couple learning-based agents with safety checks so edge cases are scored before field deployment in automotive and industrial systems.
Experience
Selected appointments
-
2024–
GenAuto AI
Executive Director
Agentic product design for autonomy; generative agents for edge-case discovery and risk assessment before release.
-
2023–24
Nuro
Senior Autonomous Vehicle Researcher
ODD-aware fault response for ML-driven, agentic driving systems.
-
2020–23
Nokia Bell Labs
Robotics Research Scientist
Functional safety and predictive control for industrial mobile robots; digital twins for manufacturing operations.
-
2016–20
Verizon
Principal Research Engineer
Edge compute and 5G for connected vehicles and robots; automation for verification and validation.
-
2013–14
KUKA Robotics
Machine Vision and Robotics Engineering Intern
Industrial automation and machine vision.
IP
Patents
- 2025 Trust through transparency: explainable social navigation for autonomous mobile robots via vision-language models
- 2024 Safe agile hazard avoidance system for autonomous vehicles
- 2024 Motion planning and control method of an overactuated four-wheel drive with constrained independent steering
- 2021 Smart IoT self-healing network
- 2019 Synchronization for battery powered IoT networks
Stack: Python, PyTorch, ROS, MATLAB/Simulink, C/C++, simulation.