Publications

Posts Coming Soon

Explore other categories in this blog or check back later.

2026

An Integrated Framework for Forecast-Driven Priority-Aware Energy Allocation in Microgrids

Presents an integrated energy management framework that combines deep learning forecasting, metaheuristic optimization, and Model Predictive Control (MPC) for priority-aware energy reallocation.

An IoT-Driven Machine Learning Framework for Real-Time Monitoring in Smart Buildings

Presents an IoT-based machine learning framework that continuously collects sensor data from smart buildings. The system enables real-time monitoring of building conditions and supports faster operational decisions.

Graph-Based Bi-Level Reinforcement Learning for Coordinated EV Fast Charging Under Distribution Constraints

Proposes a two-level reinforcement learning method built on a graph model of the distribution network to coordinate fast charging of electric vehicles. It balances charging demand while respecting grid capacity limits.

Context-Aware Functional Security for V2X-Enabled Connected Vehicles

Introduces a context-aware security approach for connected vehicles communicating over V2X links. Security responses adapt to the driving situation rather than applying fixed rules.

A Review of AI-Based Optimization and Monitoring Using Wearable Data in Smart Buildings

Reviews how data from wearable devices can feed AI models for monitoring and optimizing smart buildings. It summarizes existing methods and highlights open research gaps.

A Multi-Level Reinforcement Learning Framework for Smart Building Management

Develops a layered reinforcement learning framework where decisions are split across several control levels of a smart building. The structure improves scalability of building management.

2025

Integrative ITS Platform for Enhanced Urban Mobility in Smart Cities

Describes an integrated intelligent transportation systems platform combining traffic sensing, control and data services. The goal is smoother urban mobility in smart cities.

Addressing Urban Food Insecurity through Data-Driven and Community-Centric Smart City Frameworks

Combines data analytics with community-centred design to tackle urban food insecurity. Proposes smart city frameworks that link data, residents and local organizations.

2024

A Feedback-based Decision-Making Mechanism for Actor-Critic Deep Reinforcement Learning

Adds a feedback-driven decision mechanism to actor-critic deep reinforcement learning. The addition improves learning stability and decision quality.

OpenCyberCity Testbed's Recent Progress in Smart City Management

Reports recent progress on the OpenCyberCity scaled smart-city testbed at VCU. Covers new management capabilities added to the platform.

A machine learning-based temperature control and security protection for smart buildings

Uses machine learning both to control indoor temperature and to detect security threats in smart buildings. The two functions share the same sensing infrastructure.

A review of IoT-based smart city development and management

A broad and widely cited review of IoT technologies used in building and managing smart cities. Surveys architectures, applications and open challenges.

Demand-response prediction in smart grids using machine learning techniques

Compares machine learning models for predicting demand-response behaviour in smart grids. The predictions help operators plan load adjustments.

2023

Emulation of smart grid technologies and topologies in a small scale smart city testbed

Shows how smart grid technologies and network topologies can be emulated inside a small-scale smart city testbed. Enables low-cost experimentation with grid designs.

Optimizing smart city water distribution systems using deep reinforcement learning

Applies deep reinforcement learning to operate urban water distribution networks more efficiently. The controller learns pumping and flow policies from simulation.

Using Innovations in Data Analytics and Smart Technologies to Fight Opioid Overdose Crisis

Explores how data analytics and smart-city technologies can support responses to the opioid overdose crisis. Outlines a data-driven public health approach.

Assessing the suitability of different machine learning approaches for smart traffic mobility

Benchmarks several machine learning approaches on smart traffic mobility tasks. Identifies which model families suit which traffic problems.

Deep learning approaches for vehicle and pedestrian detection in adverse weather

Evaluates deep learning detectors for finding vehicles and pedestrians under rain, fog and other adverse weather. Highlights the accuracy drop and ways to reduce it.

Performance evaluation of a cloud-based IoT platform for smart cities: OpenCyberCity

Measures the performance of the cloud-based IoT platform behind the OpenCyberCity testbed. Reports latency and throughput characteristics.

An Automated Continuous Synthesis and Isolation for the Scalable Production of Aryl Sulfonyl Chlorides

Describes an automated continuous-flow chemical process for producing aryl sulfonyl chlorides at scale. Includes synthesis and isolation stages under automated control.

Development of the OpenCyberCity testbed: Smart city research innovation and opportunities

Presents the design and construction of the OpenCyberCity testbed. Discusses the research opportunities the facility opens for smart city work.

2022

Automated Process in Continuous Pharmaceutical Manufacturing through Real-Time Process Monitoring and Disturbance Detection Using PAT and Online First-Principle Models

An uncertainty based predictive analysis of smart water distribution system using Bayesian LSTM approach

Uses a Bayesian LSTM network to forecast behaviour of a smart water distribution system together with its uncertainty. Uncertainty estimates make the predictions more trustworthy for operators.

A Decision Support System for Cyber Physical Systems under Disruptive Events: Smart Building Application

Builds a decision support system that helps cyber-physical systems respond to disruptive events. The approach is demonstrated on a smart building.

An experiment orchestration platform to support smart city experiential learning

Presents a platform for orchestrating experiments on a smart city testbed for teaching purposes. Students can run and observe repeatable experiments.

Uncertainty estimation in power consumption of a smart home using Bayesian LSTM networks

Applies Bayesian LSTM networks to forecast smart home electricity consumption with confidence bounds. Quantified uncertainty supports better energy planning.

Incorporation of physiological features in drowsiness detection using deep neural network approach

Adds physiological signals to a deep neural network for detecting driver drowsiness. The extra features raise detection accuracy over image-only methods.

Socially-optimal auction-theoretic intersection management system

Manages autonomous vehicle traffic at intersections using an auction mechanism. The design aims for socially optimal outcomes rather than only individual gain.

Predictive control of PV/battery system under load and environmental uncertainty

Applies model predictive control to a photovoltaic and battery system facing uncertain load and weather. The controller keeps power delivery stable under those variations.

A game-theoretic model for DDoS mitigation strategies with cloud services

Models the interaction between a DDoS attacker and a cloud-based defender as a game. The results suggest cost-aware mitigation strategies.

An end to end data collection architecture for IoT devices in smart cities

Proposes a complete architecture for collecting data from IoT devices across a smart city. Covers the path from device to cloud storage and analysis.

Cyber LOPA: An integrated approach for the design of dependable and secure cyber-physical systems

Extends layer-of-protection analysis to account for cyber attacks in safety design. Gives engineers one integrated method for dependability and security.

Development of Experiential Learning Experiences for K-12 Students Focusing on Smart Cities

Describes hands-on smart city learning activities designed for K-12 students. Aims to build early interest in engineering and smart infrastructure.

2021

Enhanced systems, apparatus, and method for improved automated and autonomous operation of logistics ground support equipment

A patent application covering systems and methods for automating logistics ground support equipment. Targets safer and more autonomous ground operations.

Design and implementation of a distributed control platform for a smart building testbed

Details the design and implementation of a distributed control platform for a smart building testbed. Control tasks are spread across cooperating nodes.

A deep learning model for forecasting photovoltaic energy with uncertainties

Forecasts photovoltaic energy output with a deep learning model that also reports uncertainty. Useful for scheduling solar generation.

Towards sustainable food security: An interdisciplinary approach

Brings engineering, social science and community perspectives together on food security. Proposes an interdisciplinary research agenda.

Adaptive control for smart water distribution systems

Develops adaptive controllers for smart water distribution networks. The controller adjusts to changing demand and system conditions.

OpenCity: An open architecture testbed for smart cities

Introduces OpenCity, an open-architecture testbed for experimenting with smart city technologies. The open design lets researchers plug in new components.

Smart building management system: Performance specifications and design requirements

A highly cited paper defining performance specifications and design requirements for smart building management systems. Serves as a reference for designers.

An automatic emergency braking system for collision avoidance assist of multi-trailer vehicle based on model prediction control

Designs an automatic emergency braking system for multi-trailer vehicles using model predictive control. Helps avoid collisions with long articulated vehicles.

A novel sensitivity analysis for optimal design of superconductive fault current limiter in microgrids

Introduces a sensitivity analysis method for sizing superconducting fault current limiters in microgrids. Leads to better placement and rating choices.

Reducing Food Access Disparities: Promising Practices in Equity, Engagement, and Collaboration

Reviews promising practices for reducing disparities in food access. Focuses on equity, community engagement and cross-sector collaboration.

2020

A neural network-based model predictive control approach for buildings comfort management

Uses a neural network model inside a predictive controller to manage indoor comfort. The learned model replaces hard-to-derive physical building models.

A review of non-lane road marking detection and recognition

Surveys methods for detecting and recognizing road markings other than lane lines. Compares classical vision and learning-based techniques.

Model-based risk assessment for cyber physical systems security

Presents a model-based method for assessing security risk in cyber-physical systems. Widely cited as a systematic risk analysis framework.

An MPC-based power management of standalone DC microgrid with energy storage

Applies model predictive control to manage power in a standalone DC microgrid with storage. Balances generation, storage and load over a prediction horizon.

A CNN-based path trajectory prediction approach with safety constraints

Predicts vehicle path trajectories with a convolutional neural network while enforcing safety constraints. Aimed at safer autonomous driving decisions.

Enhanced systems, apparatus, and methods for improved automated and autonomous operation of logistics ground support equipment

US patent covering automation of logistics ground support equipment. Describes sensing, control and coordination for autonomous operation.

Cyber LOPA: a new approach for CPS safety design in the presence of cyber attacks

Preprint introducing Cyber LOPA, a safety design approach that accounts for cyber attacks. Extends classical layer-of-protection analysis.