Academic research, publications, commercial AI deployments, and engineering explorations across transport, logistics, and applied machine learning.
Doctoral thesis at the University of Exeter developing privacy-preserving federated learning frameworks for intelligent transport systems. Enables collaborative model training across distributed road infrastructure without centralising sensitive data.
Published in the IEEE Open Journal on Intelligent Transport Systems. Demonstrates how synthetic data augmentation can compensate for data scarcity in federated settings, improving traffic flow prediction accuracy across heterogeneous nodes.
Published in IEEE Sensors 2025. Develops a maximum likelihood estimation approach to IMU sensor alignment, improving accuracy of inertial measurement in connected and autonomous vehicle applications. First author: Johan Wahlstrom.
Building AI-driven decision tools for logistics and operations teams. Includes predictive stock forecasting models, purchasing optimisation systems, and end-to-end data infrastructure for clients across the UK and Europe.
MEng thesis at the University of Exeter. Designed and trained deep reinforcement learning agents to manage traffic signal control in mixed environments with connected and autonomous vehicles, reducing average journey times in simulation.