I work on electric drives. Finding faults before they spread, and keeping control when the model is wrong.

I am a PhD candidate in Electrical and Computer Engineering at the University of Alberta, in the Advanced Control Systems and Diagnosis Lab under Prof. Qing Zhao. I work on two things. Finding faults in electric drives from the signals they give off, and designing controllers that hold up when the model of the plant is wrong.

Most fault detection papers report accuracy on clean benchmarks. Real industrial data is imbalanced and noisy, and it arrives from several sensors at once. On a plant floor a false alarm often costs more than a missed fault. A lot of my work is closing that gap, using physics-informed deep learning, synthetic feature engineering, and acoustic models for anomaly detection in power drives. I also contribute to ORION, an EU and Canada project on renewable energy integration in Alberta.

Before Edmonton I did two master's degrees, the first at UET Lahore and the second at KFUPM in Dhahran, both on power and control systems. The KFUPM thesis was on model predictive torque control of induction motor drives. More recently I have been working on adversarial robustness, mainly the question of whether you can harden a network without giving up standard accuracy to do it.

Alongside the research, I served as President of the Graduate Students’ Association in 2024–2025 and as Vice President Student Services the year before. In that term we secured $8.55M in financial support for graduate students and published Under Pressure, the first comprehensive mental health report for graduate students at this university.

• Leveraging Meta-Learner-Based Stacking Framework for Stator Winding Inter-Turn Fault Detection of LSPMSMIEEE Journal of Emerging and Selected Topics in Industrial Electronics2026
• Rotor imbalance fault classification through unified feature fusion: combining statistical signal processing with adaptive deep learning featuresEngineering Applications of Artificial Intelligence, 157 · 8 citations2025
• Implications of the sensorless predictive control for line-start permanent magnet synchronous machineIEEE Transactions on Automation Science and Engineering, 21(3) · 10 citations2023
• Weighting factors optimization of model predictive torque control of induction motor using NSGA-II with TOPSIS decision makingIEEE Access, 7 · 100 citations2019

• International Student Hub

  2026

  International students at the U of A have to search across a dozen separate pages to answer urgent questions about visas, housing or mental health. This gives them one place to start.

• On-device retinal screening from a phone camera

  2026

  LoRA fine-tuning of an open-weight model so that screening suggestions can be produced locally, from non-invasive measures and a phone scan, without sending anything to a server.