Autonomous Mars Rover
A multi-subsystem rover platform combining embedded control, FPGA vision, and autonomous navigation over a networked control stack.
This project was a proof-of-concept autonomous rover designed to navigate cluttered terrain, identify points of interest, and report structured data back to a remote control application. The system combined embedded control, FPGA-based vision, motion planning, and networked telemetry in a single platform.
I primarily worked on the control, vision, and drive portions of the system. That included coordinating peripherals around an ESP32-based controller, building image-processing pipelines for obstacle detection on FPGA hardware, and integrating the routing logic used for autonomous movement.
The project was a strong systems exercise in balancing hardware, low-level software, and real-world integration constraints across a team-built platform.
The demo shows autonomous surveying and remote supervision through the control interface. The rover reports detected objects and can also be switched into manual control when needed.
The vision pipeline used FPGA image processing to detect edges and obstacles in real time, feeding those signals into the navigation and surveying stack.