Industrial Automation, AI and Research
Park AutomationIndustrial AI Research
Get information on Thomas's current skills base, industrial experience, and current research.
A full servo control system upgrade on an automated sludge press used in wastewater treatment, replacing an obsolete Siemens 611U platform with a modern S120 drive system. The project was delivered through Park Automation for a water sector integrator client.
Northern Ireland
A bespoke Automatic Licence Plate Recognition system designed and built for a waste management and land regeneration business, combining edge computing, solar power management, cloud-based plate recognition, and automated compliance reporting. No suitable off-the-shelf solution existed for the operational constraints.
Northern Ireland
A deep learning vision system built to classify date codes on packaging in a live food manufacturing environment, using a convolutional neural network running on a Raspberry Pi. Developed as an MSc project at Ulster University and designed to solve a real quality control problem that had cost the business tens of thousands of pounds.
Northern Ireland
Thomas Hadden, IEng, is an engineer and applied AI researcher based in Northern Ireland. He works at the intersection of industrial automation, process control, and artificial intelligence, with a focus on making these technologies practical, safe, and useful in real operational environments. He is an Application Engineer at Park Automation, working as a Siemens specialist across control system design, servo drive upgrades, and automation delivery for water, wastewater, and process industries. Alongside this, he is undertaking a part-time PhD at Ulster University focused on applying reinforcement learning and digital twin methods to real-world industrial process control, while Industrial Analytics & Automation (IA-2) serves as his independent research initiative outside his employed role.
Learn more
Research into detecting roller element bearing faults using wavelet decomposition and unsupervised machine learning, eliminating the need for application-specific equipment configuration. Co-authored with Dr Muhammad Usman Hadi at Ulster University and submitted to Elsevier.
Applied AI
A part-time PhD at Ulster University investigating how reinforcement learning and digital twin methodologies can be applied to real-world industrial process control, with a focus on the gap between academic theory and practical adoption in safety-critical environments.
Industrial AI
A research-led optimisation study investigating how hydraulic simulation, energy screening, and particle swarm optimisation can be combined to improve pump scheduling in water-treatment intake systems, with the aim of reducing energy use while preserving hydraulic feasibility, resilience, and practical compatibility with existing plant control.
Digital Twin Optimisation
I took a pumping model I was proud of to the operations manager of a plant and I could feel it not landing. The engineering was sound but it was not enough for buy-in. This is a look at the gap between what an engineer finds impressive and what a manager will pay for, and why being told to prove it was the most useful thing that could have happened.
2026-07-30T00:00:00Z
An optimiser can explore poor parameter combinations freely in a digital twin before any settings reach a live industrial plant, where failure can cost product, equipment, or safety.
2026-07-25T12:59:15Z
A simulation and a digital twin can be the same model. The difference is that a twin has to stay accurate about one specific plant, as it is right now. That starts with a hard baseline calibration at commissioning and continues as the plant drifts. Working through my own pumping model, this is a look at what actually separates a simulator from a twin, and why fidelity is about matching reality rather than adding detail.
2026-07-21T12:00:00Z
Assess your company's AI potential.
