Wheeled robot simulation · Demo 01
A simulation recording showing the robot's motion alongside its joint control panel.
Open demo 01 videoDr Yongxiang Lei · Singapore
From prediction
to trustworthy
decisions.
I develop interpretable, uncertainty-aware learning and control methods for renewable energy, climate forecasting, and industrial systems.
Research Scientist
SIMTech, A*STAR
Trustworthy AIProbabilistic forecastingPredictive controlDigital twins
01 / Research
Learning with purpose.
Engineering with insight.
Connecting data-driven models with the physical systems they help us understand and control.
Research focus
Forecasting that informs control
Probabilistic wave prediction accounts for uncertainty in the future excitation of a wave energy converter. Connecting those forecasts to model predictive control brings prediction and energy-conversion decisions into one framework. My broader research interests include wind-turbine pitch control and reinforcement learning.
Methods & interests
02 / Research demos
Robotics in motion.
Two simulation recordings of a wheeled robot in MuJoCo. Play either video to explore the robot's motion and joint controls.
Wheeled robot simulation · Demo 02
A second simulation recording showing another sequence of wheeled robot motion.
Open demo 02 video03 / Selected work
Research in print.
A selection of published work across climate, renewable energy, industrial modelling, and control.
2026
Control & autonomy
Robust Tube-Based Model Predictive Control for Docking Process Self-Balancing Control of a Reconfigurable Unmanned Ground Vehicle
IEEE Transactions on Automation and Science Engineering
Research focus
Robust tube-based predictive control for self-balancing during the docking of a reconfigurable unmanned ground vehicle.
2026Online 2025
Climate · Interpretability
Tem²-KAN: Data-driven temporal temperature prediction via an improved Kolmogorov–Arnold network
ISA Transactions
Research focus
A Kolmogorov–Arnold network for temperature forecasting, using learnable univariate functions to model temporal relationships.
2025
Climate · Interpretability
Climate Temporal Temperature Prediction via an Interpretable Kolmogorov–Arnold Neural Network
IEEE Transactions on Instrumentation and Measurement
Research focus
Interpretable temporal temperature prediction with a Kolmogorov–Arnold neural network.
Explore climate forecasting code2025
Industry · Uncertainty
Data-centric probabilistic temperature prediction with quantified uncertainty using Bayesian machine learning for the aluminum electrolysis process
Control Engineering Practice
Research focus
Bayesian learning for probabilistic process-temperature prediction and uncertainty quantification in aluminum electrolysis.
2024
Energy · Prediction & control
A wave forecasting method based on probabilistic diffusion LSTM network for model predictive control of wave energy converters
Applied Soft Computing
Research focus
Probabilistic diffusion and temporal modelling are connected to model predictive control for wave energy conversion.
2023
Industry · Prediction & control
Intelligent Optimal Framework for the Industrial Mining Plant-Wide Prediction Control
IEEE Transactions on Instrumentation and Measurement
Research focus
A prediction-and-control framework addressing the interactions of industrial mining processes at the plant level.
2023
Industry · Soft sensing
DualLSTM: A Novel Key-Quality Prediction for a Hierarchical Cone Thickener
Control Engineering Practice
Research focus
Temporal modelling for key-quality prediction in a hierarchical cone thickener.
2022
Industry · Self-supervised learning
A Self-Supervised Temporal Temperature Prediction Method Based on Dilated Contrastive Learning
Journal of Process Control
Research focus
Self-supervised temporal representation learning for industrial temperature prediction.
2022
Industry · Digital twins
A Digital Twin Model of Three-Dimensional Shading for Simulation of the Ironmaking Process
Machines
Research focus
Digital-twin modelling and simulation for the ironmaking process.
No matching publications
Try a different topic or search term.
Research monograph
Academic Press / Elsevier
ISBN 9780443367595Learning-Based Soft Sensing and Predictions for Process Industries
Theory, Methodology and Applications
Learning-based modelling, prediction, and soft sensing for industrial processes.
Explore the book04 / About
Across disciplines.
Grounded in engineering.
My work brings together machine learning, uncertainty quantification, and the dynamics of physical systems.
I am a Research Scientist at the Singapore Institute of Manufacturing Technology (SIMTech), A*STAR. My research spans interpretable neural networks, probabilistic forecasting, reinforcement learning, model predictive control, and digital twins.
I received my PhD in Mechanical Engineering (Control) from Politecnico di Milano in 2024. My research and teaching at the University of Warwick have focused on learning-based prediction and control for offshore renewable and sustainable energy.
Download curriculum vitaeEducation & doctoral research
- Doctoral researchUniversity of Warwick
Engineering · Prediction and control for offshore renewable energy
- 2024Politecnico di Milano
PhD · Mechanical Engineering (Control)
- 2020Central South University
MSc · Control Engineering
- 2017University of South China
BSc · Automation
Teaching
Sharing ideas through practice.
Graduate teaching at the University of Warwick in nonlinear control, reinforcement learning, and systems modelling.
- ES4G4 · Nonlinear Control and Reinforcement Learning
- ES197 · Systems Modelling, Simulation and Computation
Academic service
Contributing to the research community.
Guest editing on climate forecasting and soft sensing, and prediction and control for offshore energy.
- Applied Sciences · Climate modelling and applications
- Energy Engineering · Offshore energy prediction and control
Recognition
Selected awards.
- Best PhD Dissertation AwardPolitecnico di Milano · 2024
- Oral Presentation WinnerPostgraduate Research Showcase, University of Warwick · 2025
05 / Connect
Better questions.
Better engineering decisions.
For research collaborations in trustworthy AI, forecasting, predictive control, and digital twins.