Cornell Electric Vehicles

Electrical Team Lead, Software Team Member

2020 - present

Internationally-competing team of 40+

dedicated to designing and building fully autonomous hyper battery-efficient road vehicles.

Cornell Electric Vehicles (CEV) competes yearly in the Shell Eco-Marathon Americas competition in California, the highest echelon of international hyper fuel-efficiency competitions. Participating in a significantly more sophisticated category than Silverback Engineering, until 2021, CEV has competed in the battery-electric prototype division. For 2022 and beyond, however, CEV is transitioning into the autonomous urban-concept battery-electric category - developing a car for the fastest-growing automotive space of the decade.

Being inducted into two critical subteams working on this autonomous vehicle as a freshman is a testament to ability and passion for the project. As lead of the Electrical Team I oversee the design, manufacture, and test of all electrical and embedded systems on the vehicle. This includes automation, battery management, power converters, data acquisition, motor controller, safety, and more auxiliary systems. As part of the Software Team I work on computer vision - implementing a YOLO model to improve our car's ability to identify and track static and dynamic entities on the road.

FA 21, SP 22 Updates Coming Soon

Electrical Automation, Spring 2021:

in my first semester on CEV, I was given the opportunity to lead 'Electrical Automation', a critical project for integrating the vision and motion planning system with the drivetrain and steering system of the car.

The project, still ongoing, seeks to develop two PCBs (a main and a failsafe) to translate I2C and GPIO signals from an Nvidia Jetson into corresponding DC motor actuation. Using an iterative design approach, and prototyping using a myriad of lab tools, an ATMega chip and L298N driver design was chosen.

Having spearheaded close coordination with members on the software, mechanical, and systems subteams, the project is now onto the PCB design stage, using Altium Designer. It is expected to move into manufacture early in Fall 2021.

Electrical Automation Wiki

Software Autonomy, Spring 2021:

in my first semester on CEV, I exercised a multitude of skills in reading and implementing machine learning models from research papers. While a range of papers on convolutional neural networks were covered, significant time was spent on the ResNet and the YOLO architectures.

Sub-concepts like skip connections, bounding boxes, and scaling were among the key takeaways of the semester. Implementing models with these aided in understanding their importance to the architectures and, more generally, to CEV's dynamic object detection systems endeavours.

Through the semester I also shadowed CEV upperclassmen working towards the Shell Eco-Marathon Virtual Autonomous Programming Competition, achieving a commendable 3rd globally. Motion planning algorithms were a key element of this competition, and learning of them have inspired me to take my autonomous rover project to the next level, by adding sophisticated motion planning to its existing vision system.