Greater Toronto Area, Canada
I do large-scale ML, systems controls, and advanced signal processing for autonomous vehicles and robots.
Delivered online learning trajectory and steering control software for supercruise and hands-on lane centering.
• Developed software for control systems for automated driving features like super cruise using custom optimal algorithms developed in Matlab/Simulink, and C/C++ subsequently supported activities to deploy the C code in embedded platforms. • Developed Machine Learning software and analysis for algirthmic control and modeling, utilizing advanced fault/processes detection and signal processing. • Verified and validated performance of the controls systems through Model in Loop (MIL) simulation, Hardware in Loop (HIL) simulation, and execution and analysis of vehicle test data. • Performed Vehicle testing of developed software in production vehicles using cadence-by-cadence patches. • Conducted peer-review for software artifacts and the developed/maintained process to ensure the quality of the developed software component. • Received DFSS Green Belt, Black Belt, and 22 Record of Invention awards.
• Team leader of Research of a large-scale project on the development of advanced autonomous heavy-duty machinery utilizing state-of-the-art machine learning and path planning strategies. • Development of novel hybrid machine learning methods for condition monitoring of an electro-mechanical system.
• Increasing the tracking accuracy of an autonomous excavator to 97.4% by writing an optimized C program for a fractional-order PID controller with experimental parameter tuning using the Nelder-Mead optimization. • Lowered the cost of the DAQ system to a minimum of 200$ compared to the commercially available products in the price range of 3000-5000$, by developing customer hardware and control systems, including the electric circuits. • Upgraded the control strategy to incorporate fault detection algorithms using advanced signal processing in the form of wavelet transform and fractal dimensions. Submitted Paper. • Develop a path planning algorithm for the 7 DoF Kinova robot using particle swarm optimization in MATLAB. Submitted Paper.
• Participated in Planning and implementing the incremental improvement of the automation equipment to meet the national Industry 4.0 standards. • Put safety measures in place for production lines that eliminated active-duty accidents by employing safety redundancies. • Developed intuitive HMIs for frequently used PLCs which contributed to fast-paced control and troubleshooting.
instructor : Prof. Taghavipour
Instructor : Prof. Nahvi
• Improving the accuracy of lateral motion of a Peugeot 206 to 98.1% by developing and implementing a fuzzy model predictive controller. Published Paper. • Utilized ANFIS method to model driver's behavior and adjust the controller based on this trained model of the driver. Published Paper. • Designed a DAQ and control system using raspberry pi and OBD2 protocol interfacing with Vehicle’s ECU in python. Supervisor: Prof. Ali Ghaffari