Break

Research Interests:
My research focuses on utilizing computer vision to automate chemical processes. I am particularly interested in developing technologies that advance high-throughput experimentation, robotic platforms, process scale-up, chemical engineering, and manufacturing applications.

Current work:
In process chemistry, visual cues—such as color changes during titration, liquid levels during distillation, or solid formation in crystallization—play a crucial role in decision-making. However, current automation systems often overlook these visual signals, leaving chemists to manually handle time-consuming tasks based on these observations. My Ph.D. research addresses this gap by developing a computer vision (CV) system that enables real-time monitoring and control of chemical processes, enhancing both automation and efficiency.

My work showcases the modularity of this CV system, which can be reconfigured to automate a wide range of purification processes across different scales and throughput levels. I’ve developed three key frameworks that integrate this system into various stages of the pharmaceutical pipeline:

1. High-Throughput Experimentation:¹ In collaboration with the Aspuru-Guzik group at the University of Toronto, I designed a CV system that can monitor multiple visual cues across multiple vessels simultaneously, enabling high-throughput experimentation. I paired the software with a modular hardware system that is cost-effective and open source. This platform was applied to early-stage screening processes to optimize impurity recovery, excess reagent removal, and Grignard workups. The system automates these tasks in parallel, accelerating experimental timelines while maintaining accuracy.
2. Reactor System - Scale Up:² In collaboration with Pfizer, the CV system was scaled up and applied to automated reactor systems such as the EasyMax, allowing for vision-based, real-time adjustments to experimental conditions. This approach was employed in diverse purification processes, including liquid-liquid separation, crystallization, distillation, and titration. The system bridges high-throughput optimization to reactor-scale processes, ensuring that conditions optimized on the small scale can be seamlessly transferred to larger operations.
3. Robotic Platform for Iterative Optimization:³ In collaboration with Pfizer and Telescope Innovation Inc., I developed a CV-guided robotic platform integrated with modular hardware to enable iterative small-scale optimization, particularly for solubility and drug formulation processes. The closed-loop experiments are guided by real-time visual feedback, allowing for iterative optimization of experimental designs. This workflow was used in late-stage pharmaceutical development, leveraging data from serial experiments to drive optimization.

Through this work, I have focused on advancing autonomy in automated systems by combining visual cue monitoring (data-driven decision-making) and modular hardware for experimental workflows. The resulting self-driving lab operates across varying scales, throughput levels, and degrees of autonomy, making it adaptable to pharmaceutical pipelines at multiple stages.

Figure. A) Framework of a computer vision system that integrates real-time monitoring and feedback control for chemical processes. B) Increasing levels of autonomy achieved through advances in both software and hardware.

References:
(1) El-khawaldeh, R.; Mandal, A.; Yoshikawa, N.; Zhang, W.; Corkery, R.; Prieto, P.; Aspuru-Guzik, A.; Darvish, K.; Hein, J. E. From Eyes to Cameras: Computer Vision for High-Throughput Liquid-Liquid Separation. Device 2024, 2 (7), 100404.
(2) El-khawaldeh, R.; Guy, M.; Bork, F.; Taherimakhsousi, N.; Jones, K. N.; Hawkins, J. M.; Han, L.; Pritchard, R. P.; Cole, B. A.; Monfette, S.; Hein, J. E. Keeping an “Eye” on the Experiment: Computer Vision for Real-Time Monitoring and Control. Chem. Sci. 2024, 15 (4), 1271–1282.
(3) El-khawaldeh, R.; Hein, J. E. Balancing Act: When to Flex and When to Stay Fixed. Trends Chem. 2024, 6 (1), 1–4.

Teaching interests:
With an undergraduate degree in Biopharmaceutical Sciences – Medicinal Chemistry and a Ph.D. in Chemical Engineering in progress, my teaching interests span a range of subjects, including:
- Physical Organic Chemistry
- Thermodynamics and Kinetics
- Modeling and Simulation
- Machine Learning for Chemistry and Materials Science

I have over seven years of experience as a tutor and have served as a teaching assistant for general chemistry labs. As a first-year TA in the physical sciences, I faced the challenge of condensing vast amounts of information into digestible formats while engaging a diverse student body. To address these challenges, I highlight the real-world relevance of the course material, connecting it to various practical applications. By demonstrating how fundamental concepts apply to fields like medicine, electronics, and energy, I make the material relatable and emphasize its significance in solving real-world issues. Overall, my teaching philosophy focuses on nurturing curiosity, promoting critical thinking, and connecting course material to real-world applications, empowering students to become lifelong learners equipped to tackle challenges in any field.