Daniel Sioldea

Background

About Daniel Sioldea

I am a graduate of the Master's of Applied Science program in Electrical and Computer Engineering at McMaster University, having done my research in deep reinforcement learning for traffic light timing optimization. I had previously graduated with a Bachelor of Technology in Automation Engineering Technology from McMaster University in December 2022. My masters research project developed and validated a Duelling Double Deep-Q Network algorithm that was used to optimize traffic light timings which reduced vehicle wait times and intersection queue lengths. This model was trained and tested on the SUMO microscopic traffic simulator, where a dramatic improvement in vehicle movement was noticed. The model was then deployed onto the CARLA driving simulator where practical, user-controlled, results were obtained, noticing over 30 seconds of improvement in trip travel times. This work was done in parallel with other researches at the McMaster Automotive Resource Center, and their work on creating a robust vehicle detection and tracking model for Fisheye cameras, to present a practical and easily implementable solution to existing traffic monitoring architecture. As a Graduate Teaching Assistant at McMaster University, I supported the learning and development of undergraduate students in the Electrical and Computer Engineering department. I managed labs, and graded assignments and exams using my strong communication, and troubleshooting skills. During my undergraduate studies, I was a team member of the McMaster Engineering EcoCAR Challenge, where I contributed to the design, modelling, and testing of a hybrid vehicle. I also obtained a FANUC certification in handling tool operations and programming, gained extensive experience in 3D design using SOLIDWORKS and Autodesk Fusion 360, and designed and 3D printed numerous components for various projects.