Rigorous methods for verifying, validating, and optimizing AI in autonomous vehicles, energy systems, and critical infrastructure.
We develop mathematical tools that ensure AI systems function reliably in high-stakes applications — from safety-critical driving scenarios to mineral supply chains and energy resilience.
📍 King Fahd University of Petroleum and Minerals (KFUPM), Dhahran, Saudi Arabia
🏛️ College of Computing and Mathematics · Industrial & Systems Engineering Department
🔬 Affiliated with IRC Smart Mobility and Logistics and KFUPM-SDAIA Joint Research Center in AI
| Theme | Focus Areas |
|---|---|
| Safety & Verification | Neural network verification, adversarial robustness, deep importance sampling, runtime monitoring |
| Optimization & Decision Making | Sequential decision-making under uncertainty, reinforcement learning, Bayesian methods, safety-guaranteed autonomy |
| Sustainability & Supply Chains | AI for mineral exploration, energy resilience, critical minerals supply chains, clean energy transitions |
| Repository | Description |
|---|---|
vnvspec |
V&V-grade specifications for engineered systems — AI, optimization, simulation, and physics |
deepbullwhip |
Multi-tier supply chain bullwhip effect simulator |
notebooks |
Open notebooks from the lab — runnable in Google Colab |
Petro-HRCD-FLP |
Human-robot co-dispatch for petro-site surveillance |
- Deep Importance Sampling — safety validation in weeks rather than decades of road testing
- Adaptive Meta-Learning — out-of-distribution detection for autonomous driving (CPS Rising Star 2024)
- Diffusion-Based Failure Sampling — generating safety-critical scenarios via generative models
- Mineral-X Supply Chain Models — AI-driven optimization for critical mineral processing
50+ publications across venues including AISTATS, IEEE RA-L, IEEE T-ITS, ICRA, and CVPR.
Full list at ai-vnv.kfupm.io
We recruit postdocs, PhD students, master's students, research interns, and visiting faculty year-round.
Learn more and apply at ai-vnv.kfupm.io/apply.
