Habeeb D. Olufowobi

Research

My research focuses on advancing the security, resilience, and trustworthiness of intelligent cyber-physical systems (CPS) operating in adversarial and safety-critical environments. I develop secure and explainable AI-driven architectures for autonomous vehicles, edge intelligence, connected infrastructure, healthcare AI, and distributed cyber-physical systems. My work integrates cybersecurity, real-time systems, machine learning, and embedded systems to build intelligent systems that remain reliable under attack, uncertainty, and resource constraints.

Cyber-Physical Systems SecurityAutonomous Vehicle SecurityTrustworthy AIEdge IntelligenceExplainable AIIoT SecurityHealthcare AIReal-Time Systems

Research Thrusts

Secure and Resilient Autonomous Systems

I study how attacks against perception, communication, and control layers can compromise autonomous and connected vehicles, and I develop defenses that preserve safety and operational integrity under adversarial conditions.

  • Connected and autonomous vehicle security
  • Sensor and perception attacks
  • In-vehicle network defense
  • Runtime resilience and adversarial autonomy

Trustworthy AI for Safety-Critical CPS

I develop methods for making AI-enabled cyber-physical systems more robust, explainable, and auditable when deployed in environments where failures can affect safety, security, or public trust.

  • Adversarially robust AI/ML
  • Runtime monitoring and AI assurance
  • Explainable AI for safety-critical decisions
  • Security-aware learning systems

Secure Edge Intelligence and Distributed CPS

I investigate secure and efficient intelligence at the edge, including resource-constrained AI, federated learning, IoT/IIoT security, and distributed cyber-physical infrastructures.

  • Edge AI and resource-constrained reasoning
  • IoT and industrial IoT security
  • Federated and privacy-preserving learning
  • Distributed intelligent infrastructure

Explainable and Equitable AI for Healthcare

I apply trustworthy AI principles to biomedical and healthcare problems, with emphasis on explainability, fairness, robustness, and health-equity-aware machine learning.

  • Chagas disease diagnosis and prediction
  • Medical imaging AI and model explainability
  • Fairness and bias mitigation in clinical AI
  • Responsible deployment of healthcare AI/ML

Selected Ongoing Projects

Vehicle Security

Securing Connected Autonomous Vehicle Stack Against Adversarial Input

This project investigates cross-layer security risks that emerge when adversarial inputs propagate from autonomy functions, such as perception and decision-making, into low-level vehicle control and communication systems.

Related Publications

  • VehicleSec 2025 “Beyond the Glow: Understanding Luminescent Marker Behavior Against Autonomous Vehicle Perception Systems.”
  • IEEE T-ITS 2024 “From Weeping to Wailing: A Transitive Stealthy Bus-Off Attack.”
  • VehicleSec 2024 “AutoWatch: Learning Driver Behavior with Graphs for Auto Theft Detection and Situational Awareness.”
  • IEEE TVT 2019 “SAIDuCANT: Specification-based Automotive Intrusion Detection using CAN Timing.”
XAI in Healthcare

Explainable AI for Chagas Disease Diagnosis and Prediction

This project develops ethical, explainable, and trustworthy AI/ML models for diagnosis and prediction of heart disease outcomes in Chagas disease, with attention to transparency, robustness, and health equity.

Related Publications

  • WACV 2026 “MorphXAI: An Explainable Framework for Morphological Analysis of Parasites in Blood Smear Images.”
  • BHI 2025 “Beyond Detection: Comparative Explainability Study on Trypanosoma cruzi Using CAMs and DETR Attention.”
  • AIM-AHEAD 2025 “Beyond Detection: A Trustworthy and Explainable AI Framework for Chagas Disease Using YOLOv8 and DINO-DETR.”
  • AIM-AHEAD 2024 “Improving Health Equity in Algorithmic Decision-Making Using Explainable AI Techniques.”
Trustworthy Edge AI

Secure Edge Intelligence for CPS Platforms

This research explores secure and analyzable AI reasoning at the edge for resource-constrained CPS platforms, including IoT, autonomous systems, and distributed sensing infrastructures.

Related Publications

  • IEEE IoTJ 2024 “D-NDNoT: Deterministic Named Data Networking for Time-Sensitive IoT Applications.”
  • IEEE Access 2024 “Simulating Load Sharing for Resource Constrained Devices.”
  • IEEE EDGE 2023 “FedCime: An Efficient Federated Learning Approach for Clients in Mobile Edge Computing.”
  • IEEE WF-IoT 2021 “Dynamic Load Sharing in Memory Constrained Devices: A Survey.”
AI + Transportation

Privacy-Preserving and Resilient Intelligent Transportation Systems

This work studies secure data sharing, privacy-preserving learning, and resilient communication architectures for intelligent transportation systems and Internet of Vehicles environments.

Related Publications

  • OJ-COMS 2026 “Digital Twin–Guided AI Path Planning for Connectivity-Aware Mobility.”
  • ICC 2026 “CTMap: LLM-Enabled Connectivity-Aware Path Planning in Millimeter-Wave Digital Twin Networks.”
  • ESANN 2026 “Privacy-preserving Intrusion Detection System for Internet of Vehicles Using Split Learning.”
  • BDCAT 2023 “Privacy-Preserving Intrusion Detection System for Internet of Vehicles Using Split Learning.”
AI Security

Robust Vision-Language Models for Autonomous Systems

This project investigates vulnerabilities and defenses for multimodal AI systems used in autonomous navigation, including perception attacks, federated personalization, and robust inference.

Related Publications

  • ICML 2026 “Revisiting Asymmetries in Black-box Link Stealing against Graph Neural Networks.”
  • PETS 2026 “Unveiling Graph Copycats: Inference Attacks with Student Models.”
  • IV 2026 “Toward Inherently Robust VLMs Against Visual Perception Attacks.”
  • ECAI 2025 “FedVLM: Scalable Personalized Vision-Language Models through Federated Learning.”
Cyber-Physical Defense

Runtime Security for Embedded and Real-Time Systems

This research develops intrusion detection, prevention, and recovery mechanisms for embedded and real-time platforms, including automotive networks and industrial control systems.

Related Publications

  • DSN 2025 “MichiCAN: Spoofing and Denial-of-Service Protection Using Integrated CAN Controllers.”
  • IECON 2024 “MCFICS: Model-based Coverage-guided Fuzzing for Industrial Control System Protocol Implementations.”
  • IEEE TVT 2023 “CANASTA: Controller Area Network Authentication Schedulability Timing Analysis.”
  • RTSS 2021 “Vulnerability of Controller Area Network to Schedule-Based Attacks.”

Funding and Research Impact

$598,895

NSF CAREER Award

Supporting research on securing connected autonomous vehicle stacks against adversarial input.

$250,000

NIH AIM-AHEAD Award

Supporting explainable AI/ML models for Chagas disease diagnosis and prediction.

$50,000

NIH AIM-AHEAD Fellowship

Supporting health-equity-focused research on explainable AI for algorithmic decision-making.

$25,000

Google-CAHSI Fellowship

Supporting cybersecurity curriculum development and training at Hispanic-Serving Institutions.

Research Infrastructure

The Cyber-physical Systems Security Lab supports research in AI security, autonomous systems, embedded systems, healthcare AI, and large-scale machine learning experimentation. The lab includes multi-GPU compute resources and dedicated infrastructure for cyber-physical systems security, trustworthy AI, and edge intelligence research.

Research Sponsors