Context-driven Deep Learning Forecasting for Wastewater Treatment Plants
Sikder, M. N. K., & Batarseh, F. A. (2025). "Context-driven Deep Learning Forecasting for Wastewater Treatment Plants." ACM Transactions on Cyber-Physical Systems.
CV
Download the full CV: PDF · short version
Publications appear under M. N. K. Sikder; pre-2016 work under N. K. Sikder.
Professional Summary
Research Assistant Professor focused on Generative AI security, cybersecurity, critical infrastructure protection, and trustworthy embedded AI. Previously a Presidential Postdoctoral Fellow specializing in secure, privacy-aware, and explainable AI for critical infrastructure. My research spans hybrid AI — including LLMs, agentic AI, computer vision, and time-series forecasting — to secure AI-enabled systems, support anomaly detection, and enable robust decision-making in cyber-physical environments, with interdisciplinary publications and contributions to NSF proposal development.
Areas of Expertise
- Critical Infrastructure and Control Systems. SCADA and industrial control system (ICS) security; water and wastewater treatment systems; operational technology (OT) monitoring; cyber-physical system resilience.
- Cybersecurity. Intrusion detection and automated response; IoT and cross-layer (network/host) threat detection; adversarial machine learning; threat-intelligence grounding (MITRE ATT&CK, CVE/NVD).
- Artificial Intelligence. Security and trustworthiness of large language models and agentic AI systems; AI assurance, validation, and explainability; anomaly detection; time-series forecasting; generative modeling (GANs, diffusion models).
- Digital Media Forensics. Detection and source attribution of synthetic and adversarially manipulated imagery.
- Telecommunications. LTE radio access network rollout, operations, and monitoring automation.
Academic Appointments
- Research Assistant Professor, Old Dominion University, Norfolk, VA — December 2025 to present
- Presidential Postdoctoral Fellow, Virginia Tech, Arlington, VA — January 2025 to December 2025
- Graduate Research Assistant, Virginia Tech, Arlington, VA — August 2019 to December 2024
Education
- Ph.D., Computer Engineering, Virginia Polytechnic Institute and State University, Arlington, VA — August 2019 to December 2024
- Dissertation: AI Methods for Anomaly Detection in Cyber-Physical Systems: With Application to Water and Agriculture. Chair: Feras A. Batarseh. Link
- M.S., Computer Engineering, Virginia Polytechnic Institute and State University, Falls Church, VA — August 2019 to May 2022
- GPA: 3.67/4.0
- B.S., Electrical and Electronic Engineering, Bangladesh University of Engineering and Technology (BUET), Dhaka, Bangladesh — May 2010 to September 2015
- GPA: 3.65/4.0
Honors, Awards, and Fellowships
- Presidential Postdoctoral Fellow (competitive university-wide fellowship), Virginia Tech, 2025
- Winner, 2022 Intelligent Water Systems Challenge, The Water Research Foundation
- 2nd best team, line following robot contest, BUET, 2014
Research Interests
Security and Trustworthiness of LLMs and Agentic AI; Cybersecurity; Trustworthy and Secure Embedded AI for Cyber-Physical Systems; Anomaly Detection and AI Assurance.
Publications
Assessing the Fidelity and Utility of Water Systems Data Using Generative Adversarial Networks: A Technical Review
Sikder, M. N. K., Wang, Y., & Batarseh, F. A. (2025). "Assessing the Fidelity and Utility of Water Systems Data Using Generative Adversarial Networks: A Technical Review." IEEE Access.
AI Methods for Anomaly Detection in Cyber-Physical Systems: With Application to Water and Agriculture
Sikder, M. N. K. (2024). "AI Methods for Anomaly Detection in Cyber-Physical Systems: With Application to Water and Agriculture." Ph.D. dissertation, Virginia Polytechnic Institute and State University.
P2O: AI-Driven Framework for Managing and Securing Wastewater Treatment Plants
Kulkarni, A., Yardimci, M., Kabir Sikder, M. N., & Batarseh, F. A. (2023). "P2O: AI-Driven Framework for Managing and Securing Wastewater Treatment Plants." Journal of Environmental Engineering, 149(9), 04023045.
The application of artificial intelligence assurance in precision farming and agricultural economics
Williams, M. J., Sikder, M. N. K., Wang, P., Gorentala, N., Gurrapu, S., & Batarseh, F. A. (2023). "The application of artificial intelligence assurance in precision farming and agricultural economics." In AI Assurance: Towards Valid, Explainable, Fair, and Ethical AI, 501–529. Academic Press.
Outlier detection using AI: a survey
Sikder, M. N. K., & Batarseh, F. A. (2023). "Outlier detection using AI: a survey." In AI Assurance: Towards Valid, Explainable, Fair, and Ethical AI, 231–291. Academic Press.
Deep H2O: Cyber attacks detection in water distribution systems using deep learning
Sikder, M. N. K., Nguyen, M. B., Elliott, E. D., & Batarseh, F. A. (2023). "Deep H2O: Cyber attacks detection in water distribution systems using deep learning." Journal of Water Process Engineering, 52, 103568.
AI Assurance: Towards Trustworthy, Explainable, Safe, and Ethical AI
(2023). Academic Press.
Model-Agnostic Scoring Methods for Artificial Intelligence Assurance
Sikder, M. N. K., Batarseh, F. A., Wang, P., & Gorentala, N. (2022). "Model-Agnostic Scoring Methods for Artificial Intelligence Assurance." 2022 IEEE 29th Annual Software Technology Conference (STC), 9–18.
Time-Series Generative Adversarial Networks for Cyber-Physical Systems
Nguyen, M. B. T., Sikder, M. N. K., & Wang, C. (2022). "Time-Series Generative Adversarial Networks for Cyber-Physical Systems." Technical report and open-source release.
Realtime Management of Wastewater Treatment Plants Using AI
Batarseh, F. A., Yardimci, M. O., Suzuki, R., Sikder, M. N. K., Wang, Z., & Mao, W. (2022). "Realtime Management of Wastewater Treatment Plants Using AI." Virginia Tech & DC Water.
DeepAg: Deep Learning Approach for Measuring the Effects of Outlier Events on Agricultural Production and Policy
Gurrapu, S., Batarseh, F. A., Wang, P., Sikder, M. N. K., Gorentala, N., & Gopinath, M. (2021). "DeepAg: Deep Learning Approach for Measuring the Effects of Outlier Events on Agricultural Production and Policy." IEEE Symposium Series on Computational Intelligence (SSCI), 1–8, Orlando, FL.
Quantification of Peak Demand Reduction Potential in Commercial Buildings due to HVAC Set Point and Brightness Adjustment
Usman, M. U., Haque, A., Sikder, M. N. K., Cai, M., Bradley, S. R., Pandey, S., Kliros, C., & Zhang, L. (2021). "Quantification of Peak Demand Reduction Potential in Commercial Buildings due to HVAC Set Point and Brightness Adjustment." 2021 IEEE Power & Energy Society General Meeting (PESGM), 1–6.
Applications of Machine Learning for Precision Agriculture and Smart Farming
Gurrapu, S., Sikder, N., Wang, P., Gorentala, N., Williams, M., & Batarseh, F. A. (2021). "Applications of Machine Learning for Precision Agriculture and Smart Farming." The International FLAIRS Conference Proceedings, 34.
Implementation of microcontroller based Maximum Power Point Tracker (MPPT) using SEPIC converter
Chakma, S., Sikder, N. K., Khan, S. I., & Akhter, S. (2015). "Implementation of microcontroller based Maximum Power Point Tracker (MPPT) using SEPIC converter." 2015 IEEE International WIE Conference on Electrical and Computer Engineering (WIECON-ECE), 374–377.
Research Funding and Grant Activity
Awarded
- Commonwealth Cyber Initiative (CCI). Agentic AI for Scalable and Cost-Effective Protection of Critical Infrastructure. Total award: $100,000; period: 1 year. Role: Co-PI / Senior Personnel. PI: Daniel Takabi; Co-PI: Murat Kantarcioglu. Institutions: Old Dominion University, Virginia Tech.
Submitted / Under Review
- National Science Foundation (Cyber-Physical Systems Program). Securing Water and Agricultural Cyber-Physical Systems Using Foundational AI. Total budget: $1,000,000; duration: 4 years. Role: Senior Personnel and Methodology Lead. PI: Feras A. Batarseh; Co-PIs: Manish Bansal, Jonathan Czuba, Dong Ha, Abhilash Chandel, Azahar Ali. Institution: Virginia Tech.
Planned / Targeted Proposals (PI or Co-PI)
- NSF Secure and Trustworthy Cyberspace (SaTC 2.0). Secure and trustworthy generative-AI systems for cyber-physical and critical-infrastructure security. Target submission 2026.
- NSF Cyber-Physical Systems (CPS). AI-enabled security, privacy, and resilience for networked and embedded cyber-physical systems. Target submission 2026.
- DARPA (I2O) Office-Wide BAA. Robust and secure agentic-AI and generative-AI systems for cyber operations and critical-infrastructure defense. Rolling through 2026.
- Office of Naval Research (ONR) Long-Range BAA. Trustworthy AI and secure machine learning for resilient cyber-physical and autonomous systems. Rolling through 2026.
- National Lab Partnerships (e.g., PNNL / DOE). AI-accelerated adversary emulation and cyber-defense for water, energy, and industrial control systems.
Research Experience
Research Assistant Professor — Old Dominion University
- Co-Investigator on a CCI-funded program developing agentic AI for scalable, cost-constrained protection of critical infrastructure, spanning threat detection, automated response, and operational deployment cost.
- Designed and implemented H-CLAIR, a two-tier cross-layer intrusion detection and response framework pairing an always-on lightweight ensemble (isolation forest, random forest, autoencoder, TabNet, graph attention networks) fused by weight-of-evidence log-odds with an LLM verifier grounded in MITRE ATT&CK and CVE/NVD, and a constrained-MDP response agent over a bounded action space. Evaluated across network and host layers on ToN_IoT, SWaT, and WADI.
- Built a production-grade agentic anomaly detection pipeline for multilayer network data on the Wahab HPC cluster (SLURM scheduling, H100 GPUs, containerized inference, retrieval-augmented multi-model ensemble).
- Improved graph deviation network (GDN) anomaly detection on industrial control system benchmarks via orthogonal initialization, a frozen reconstruction decoder, and model ensembling.
- Authored or co-authored 2 federal and foundation proposals, including NSF SaTC and Schmidt Sciences submissions.
- Supervising 2 graduate students.
Presidential Postdoctoral Fellow — Virginia Tech, Commonwealth Cyber Initiative (CCI)
- Led design and prototyping of LLM- and computer-vision-based anomaly detection systems for critical infrastructure, including benchmarked evaluation harnesses comparing classical and deep learning approaches to reduce false alerts.
- Developed production-ready trustworthy-AI components: data-quality validation, retraining triggers, model cards, and containerized inference pipelines, with interdisciplinary engineering and operations teams.
- Senior Personnel and methodology lead on a large-scale NSF CPS proposal, contributing system architecture, experimental design, and evaluation frameworks.
Graduate Research Assistant — Virginia Tech, Commonwealth Cyber Initiative (CCI)
- Developed advanced AI models (High Confidence AutoEncoders, GANs) for real-time cyber-physical threat detection in water supply systems.
- Built a context-aware forecasting framework (Temporal Fusion Transformer) integrating external data for improved water systems prediction.
- Developed AI-based decision support and anomaly detection for SCADA using deep recurrent models (LSTM/GRU).
- Applied isolation-forest-based modeling for agricultural production forecasting and policy analysis.
- Contributed to AI assurance methods for fairness, security, and explainability.
Graduate Research Assistant — Virginia Tech, Advanced Research Institute
- Developed data-driven energy-efficient building models using real-time data to support HVAC optimization.
- Performed demand-response analysis for appliance scheduling and AI-based policy recommendations.
Industry and Professional Experience
AI Systems Lead — DC Water Operational Analytics, Washington, DC (2023–2025)
- Led development, deployment, and validation of AI-driven embedded systems for real-time wastewater tunnel monitoring and anomaly detection at one of the largest municipal water utilities in the United States.
- Delivered production components in live operational use: data-quality validation, retraining triggers, model cards, and containerized inference pipelines, with utility engineering and operations staff.
- Benchmarked deep learning approaches against classical baselines to reduce false alert rates supporting operator decision-making.
BEM Controls LLC, McLean, VA — Graduate Research Intern (May 2020 – August 2020)
- Developed and tested enterprise-level smart grid software (BEMOSS) for device-level load control across heterogeneous communication protocols and demand response estimation.
Grameenphone Ltd., Dhaka, Bangladesh — System Engineer (October 2015 – July 2019)
- Led LTE network rollout execution in Dhaka city; recognized for rapid rollout performance.
- Designed and maintained a Telegram bot Android application for network monitoring and maintenance.
- Developed protection systems and operational tools (billing automation, fuel generator controller) reducing operational expenditure.
Teaching
Teaching Interests: Trustworthy and Secure AI; Cybersecurity and Critical Infrastructure; Machine Learning for Cyber-Physical Systems; Generative AI and LLM Security; Applied Deep Learning (Vision and Time Series); AI Assurance and Anomaly Detection.
Advising and Mentoring
- Trey Ward (M.S. student, Virginia Tech). Co-mentored for a journal manuscript in preparation on secure and trustworthy computer vision for agricultural cyber-physical systems. Supervised problem formulation, experimental design, model development, and manuscript preparation.
- Shubham Deshmukh (M.S. student, Virginia Tech). Co-mentored on the same project, contributing dataset generation, adversarial and generative modeling pipelines, and evaluation of detection and attribution models.
- Research project: Unified Detection and Attribution of Synthetic and Adversarial Images in Agricultural Cyber-Physical Systems. Designed a multi-generator evaluation framework using GANs (StyleGAN2, StyleGAN3, R3GAN) and diffusion models (InstructPix2Pix, BLIP-Diffusion, Dreamshaper-8) across apple, maize, and tomato, guiding use of CNN and Vision-Transformer backbones (EfficientNet-B0, ResNet-50, CLIP) for health-state classification, source detection, and generator attribution.
- Trained students in reproducible ML pipelines, dataset curation, cross-model evaluation, and scientific writing.
Professional Service
Proposal Review and Panel Service
- Commonwealth Cyber Initiative (CCI), Virginia — External Reviewer (2025). Reviewed and formally scored proposals for the CCI Request for Proposals: Cybersecurity for Critical Infrastructure.
Peer Review and Editorial Service
- Environmental Monitoring and Assessment (Springer Nature), 2026
- Journal of the ASABE (American Society of Agricultural and Biological Engineers), 2025
- Journal of Computer Virology and Hacking Techniques (Springer Nature), 2025
- Scientific Reports (Nature Portfolio), 2025
- International Journal of Data Science and Analytics (Springer Nature), 2025
- Environmental Health (Springer Nature), 2025
- SN Computer Science (Springer Nature), 2025
- Neural Computing and Applications (Springer Nature), 2025
- Journal of Cybersecurity and Privacy (MDPI), 2025
Leadership and Outreach
- AI Systems Lead, DC Water Operational Analytics, 2023–2025
- Volunteer, IEEE Innovation Smart Grid Technologies (ISGT), North America, 2020
- Volunteer, AEE World Energy Conference and Expo, Washington, DC, 2019
- Coordinator, Inter-university Project Show and Departmental Festival, 2014
- Organizing Member, PES 2014 Inter-university Robotics Competition, BUET Energy Club
Talks and Presentations
- SDSS 2022 — Model-Agnostic Scoring Methods for Artificial Intelligence Assurance, Symposium on Data Science and Statistics, June 2022. Abstract
Open-Source Software and Datasets
- Context-Driven Forecasting (dataset + code). Dataset and reproducible software package supporting the context-driven forecasting study: time-series data processing, modeling, and evaluation.
- AgriVision Synthetic/Adversarial Image Detection (dataset + framework). Curated dataset for synthetic/adversarial image detection in agriculture, with a benchmark framework for detection and attribution experiments.
Projects
Technical Skills
- Programming: Python, C++, SQL
- ML/AI: PyTorch, TensorFlow, Keras; Hugging Face Transformers; Ray RLlib
- Data: Pandas, NumPy, Scikit-learn; Apache Spark
- DevOps/MLOps: Docker, Kubernetes, Jenkins, Apache Airflow; MLflow; Weights & Biases; Terraform
- Cloud/Edge: AWS (SageMaker), Azure, Google Cloud; TensorFlow Lite; AWS IoT
- Security: OpenSSL, JWT, PyCrypto
Professional Memberships
- Institute of Electrical and Electronics Engineers (IEEE) — Member; IEEE Computer Society
- Association for Computing Machinery (ACM) — Member
Media Coverage
- Intelligent Water Systems Challenge winner coverage, The Water Research Foundation
- Virginia Tech: AI solution implementation at DC Water
- Commonwealth Cyber Initiative: protecting water systems from cyber threats
References
Available on request.