Recruitment
We recruit year-round for postdocs, MASc and PhD students, visiting students and undergraduate students. Please see our Opportunities page for more information.
Research Overview
We develop AI and machine learning methods for industrial process control, with research informed by collaborations with industry partners across refining, pulp and paper, mining, and pharmaceuticals.
Our work spans both theory and application: we develop novel algorithms and computational tools to solve industrial process control problems, while also exploring fundamental problems in control theory and machine learning.
What We Do
Intelligent Control
We develop reinforcement learning and model predictive control methods that can adapt to changing process conditions without manual retuning. Our work on deep RL, meta-RL, and goal-conditioned learning enables controllers that learn from data while providing stability and safety guarantees.
Digital Twins & Modeling
We build large-scale dynamic models and hybrid physics-ML models that serve as digital twins for industrial processes. These models combine first-principles knowledge with data-driven learning to capture complex, nonlinear process behavior for simulation, optimization, and control.
Fault Detection & Diagnostics
We develop methods to detect anomalies, identify root causes, and diagnose faults in complex industrial systems. Our approaches use representation learning, contrastive methods, and interpretable ML to provide actionable insights from heterogeneous process data.
Predictive Analytics
We create soft sensors, forecasting models, and visualization tools that extract insights from large volumes of industrial data. Our methods handle multirate, multivariate time series and provide reliable predictions for process monitoring and decision support.
Where We Apply It
Energy
Oil and gas refining, battery management systems, power systems, and renewable energy. We work on real-time optimization, state estimation for lithium-ion batteries, and carbon tracking in co-processing operations.
- Oil & Gas Refining
- Battery Management Systems (Wetton Group, Cao Group)
- Power Systems (BCHydro)
- Renewables & Carbon Capture
Natural Resources
Mining operations, pulp and paper manufacturing, and biomass processing. Our work includes fault detection in mineral processing, quality control in paper machines, and process optimization.
- Mining (BBA, Teck, Baldwin Group)
- Pulp & Paper (Canfor, Spartan Controls)
- Biomass (Sokhansanj Group)
Manufacturing
Refining, chemical processing, and process control systems. We collaborate with industry partners on advanced process control, soft sensors, and controller performance monitoring.
- Refining (Parkland)
- Process Control & Optimization (Honeywell, Loewen Group)
Biotechnology & Pharma
Pharmaceutical manufacturing, bioprocess optimization, and experiment design for cell cultures. We develop models and control strategies for complex biological systems.
- Pharmaceuticals (Amgen)
- Bioengineering (Piret Group)
Material Discovery
Machine learning for accelerated discovery of materials for carbon capture and other applications. We use Bayesian optimization and multi-fidelity methods for high-throughput screening.
- CO2 Capture Materials (Ellis Group)
- Biomass-Derived Activated Carbons
Health Analytics
Wearable sensors, patient monitoring, and clinical decision support. We develop algorithms for continuous health monitoring and triage assistance.
- Patient Monitoring Systems
- Clinical Decision Support (Ho Group)
Publications
Please see our publications list for detailed information on our research outputs, including work on:
- Reinforcement Learning for Process Control
- Digital Twin Development
- Soft Sensors and State Estimation
- Fault Detection and Diagnosis
- Battery Modeling and Management
- Process Optimization