About the Role
You'll develop the team's internal computer vision software and get it running on real, resource-constrained hardware. Expect to work with OpenCV and YOLO-style detection, then optimise for edge deployment: TinyML on STM32 and RP-family chipsets, working within tight compute, memory, and power budgets. This is the challenging, rewarding side of ML: not just training a model, but making it run efficiently on the hardware that actually flies. You'll learn embedded optimisation, model deployment, and the practical craft of edge AI. Languages you'll use include C++, C, Arduino, and Python: and we'll teach you what you don't already know.
Requirements
Passion for machine learning, computer vision, and squeezing performance out of small hardware.
Programming experience in C++, C, Arduino, or Python is helpful but will be taught if you don't have it.
Curiosity about running models on constrained, low-end hardware.
About the Team
DesignBuildFly at MPSTME is a student-led aerospace design team founded in July 2025 at the Mukesh Patel School of Technology Management and Engineering, Mumbai. We design, fabricate, and test radio-controlled aircraft to compete in the AIAA Design/Build/Fly competition, working to precise mission requirements against significantly larger and better-funded programmes. In our very first year we placed 20th in Round 1 of the 2025–26 competition, and our current challenge is a banner-towing bush plane built to execute charter-flight ground operations and aerial banner deployment. Our technical philosophy centres on advanced composites carbon-fibre-reinforced polymers (CFRP) for primary structures integrated with additive manufacturing to hit exceptional strength-to-weight ratios. Every member receives structured training in flight dynamics, stability analysis, and performance prediction, and we're expanding into CubeSats too. No prior experience is expected from first-years, we teach what you don't yet know. What matters is that you're curious, committed, and want to build real aircraft that fly.