Applied AI Research in Production and Energy
Dr Peizhi Shi and colleagues have been doing research on how applied AI and machine learning can support intelligent manufacturing, energy forecasting, and operational decision-making.
Across manufacturing and energy contexts, organisations increasingly need to interpret complex technical data, translate it into meaningful operational knowledge, forecast changing demand, and make decisions under uncertainty. AI offers new ways to recognise structures and patterns in these systems, from identifying manufacturing features in digital product models and linking them to process planning, to forecasting electricity demand across different system scales and operating conditions.
These recent publications reflect this broader research agenda. Collectively, the work goes beyond applying AI to individual technical tasks; it examines how AI methods can be developed, organised, and evaluated for complex operational contexts. Across the three papers, this includes how machine learning can extract meaningful production knowledge from manufacturing data, how forecasting models can adapt in real time as demand patterns change, and how machine learning methods can be systematically understood across different peak demand forecasting settings. The three studies connect methodological insights in AI with intelligent production, process planning, energy management, operational decision-making, and sustainable resource use.
The three published papers are:
- Machine learning in feature recognition for manufacturing: taxonomy, analytical review, comparisons, trends, challenges, and outlook.
- Multiscale and Real-Time Load Forecasting: A Universal Online Functional Data Analysis Framework.
- Machine learning in peak demand forecasting: foundations, trends, and insights.
Together, these studies contribute to both AI methodology and real-world operational practice. They address technical challenges in manufacturing and energy systems while also speaking to wider questions in operations and analytics, particularly how AI methods can be designed, organised, and evaluated to support practical decision-making in industry.
