A heterogeneous GNN + scalable transformer with physics-informed positional encodings for ACOPF — up to 5,000× faster than interior point solvers.
Link to interactive app
Published in International Journal of Computing and Digital Systems, 2021
A computer vision system for recognizing human posture in surveillance settings.
Recommended citation: O. F. Arowolo, E. O. Arogunjo, D. G. Owolabi, and E. D. Markus (2021). "Development of A Human Posture Recognition System for Surveillance Application." International Journal of Computing and Digital Systems. 10(1), 1191–1197.
Download Paper
Published in arXiv preprint, 2022
Data augmentation techniques for satellite imagery in computer vision pipelines.
Recommended citation: O. Adedeji, P. Owoade, O. Ajayi, and O. Arowolo (2022). "Image Augmentation for Satellite Images." arXiv:2207.14580.
Download Paper
Published in IEEE Transactions on Sustainable Energy, 2025
Machine learning accelerates ADMM-based decomposition for stochastic power system scheduling with numerous scenarios.
Recommended citation: A. Rajaei, O. Arowolo, J.L. Cremer (2025). "Learning-Accelerated ADMM for Stochastic Power System Scheduling With Numerous Scenarios." IEEE Transactions on Sustainable Energy.
Download Paper
Published in Sustainable Energy, Grids and Networks, 2025
How well do machine learning surrogates for power system time-domain simulation extrapolate beyond their training data?
Recommended citation: O. Arowolo, J. Stiasny, J.L. Cremer (2025). "Exploring extrapolation of machine learning models for power system time domain simulation." Sustainable Energy, Grids and Networks.
Download Paper
Published in IEEE PES Innovative Smart Grid Technologies Conference Europe (ISGT Europe), 2025
Spatio-temporal graph neural networks for the multi-period optimal power flow problem.
Recommended citation: A. Rajaei, O. Arowolo, J.L. Cremer (2025). "Spatio-Temporal Graph Neural Networks for Multi-Period Optimal Power Flow." 2025 IEEE PES ISGT Europe.
Download Paper
Published in Energy and AI, 2026
Graph neural networks that generalize across grid topologies and operating conditions for AC optimal power flow.
Recommended citation: O. Arowolo, J.L. Cremer (2026). "Towards generalization of graph neural networks for AC optimal power flow." Energy and AI.
Download Paper
Published in SSRN preprint, 2026
A tabular foundation model approach to data-driven dynamic security assessment of power systems.
Recommended citation: O. Arowolo, M. Yang, J.L. Cremer (2026). "Revisiting data-driven dynamic security assessment with a tabular foundation model." SSRN.
Download Paper
Published:
Abstract: Larval zebrafish is widely used in neuroscience research because it has a small transparent nervous system, allowing for cellular resolution examination of neural dynamics throughout its brain under a microscope. Recent advances in neural imaging have developed a microscopy technique that tracks the motion of freely swimming larval zebrafish and keeps its brain within the camera’s field of view. This technique utilizes model predictive control (MPC), which depends on the accurate prediction of the animal’s future position. However, larval zebrafish moves in discrete bouts making sudden, high-velocity movements that are difficult to predict. Without predictive control, tracking errors greater than 100 µm occur about 9% of the total time and 49% of the time when the fish is in motion. To reduce this error, we use a probabilistic gradient boosting prediction framework to determine the most likely region the fish can be found in the next 6 camera frames given a sequence of the fish’s coordinates and heading vector for the previous 10 frames. We use these predictions to plan the path of the tracking microscope. Compared to without prediction, we reduced the number of frames with tracking errors above 100 µm by an average of 40%. Our results demonstrate that, given a short history of previous positions, Machine Learning (ML) techniques can be used to predict the future trajectory of larval zebrafish by uncovering informative patterns in zebrafish locomotion. Watch the talk.
Published:
Abstract: Time domain simulation (TDS) is an important tool for assessing power system security under various disturbances. However, its computational cost limits the number of disturbances that can be assessed. The need for fast assessment of numerous disturbances has increased with the rapid integration of renewable energy sources. Machine learning (ML) methods have been explored to accelerate power system TDS, but these methods are studied in interpolation scenarios, where they predict outputs for inputs within the training data distribution. This work uses a state-of-the-art ML model to explore the extrapolation behaviour of ML models for TDS. First, we highlight the importance of ML models’ extrapolation capacity for fast assessment of numerous diverse disturbances. Next, we demonstrate that extrapolation for discrete disturbances is more challenging than for continuous disturbances. Subsequently, we investigate how transfer learning (TL) may be used to improve the performance of ML models in TDS extrapolation scenarios. Finally, we outline the limitations of TL for power system TDS and suggest alternative approaches for developing ML models with better extrapolation performance in TDS applications. Read the paper.
Teaching assistant, Carnegie Mellon University Africa, 2022
I was a teaching assistant for the Photovoltaic Systems Engineering course (ECE 18-883) at Carnegie Mellon University. ECE 18-883 is a Master level course for the Electrical and Computer Science program at the College of Engineering. The course introduces students to assessment of solar resources, analysis of solar panel rating, output and efficiency, and the design of off-grid PV systems. I did the following:
Project supervision, Delft University of Technology, 2023
I was a project supervisor for the course Interdisciplinary Advanced Artificial Intelligence Project (IAAIP CS520200). IAAIP CS520200 is a Master level course for computer science students at the Electrical Engineering, Mathematics and Computer Science faculty. Students carried out research projects applying artificial intelligence to climate-related problems in various departments. I supervised a team of 4 students who researched the use of graph neural networks for power flow analysis.
Teaching assistant, Delft University of Technology, 2025
I was a teaching assistant for the Machine Learning Workflows for Digital Energy Systems course (SET 3125) at TU Delft. SET 3125 is a Master level course for the Sustainable Energy Technology Master program at Electrical Engineering, Mathematics and Computer Science Faculty. The course introduces students to statistical learning concepts, supervised regression, and training neural networks. The students also learn designing machine learning workflows, physics informed learning for energy applications and learning with inductive biases in energy grids. I did the following: