Mars Rover — University Rover Challenge (URC)
The University Rover Challenge is a premier international robotics competition for university teams to design and field a Mars rover prototype. Teams complete missions such as autonomous navigation, equipment servicing (dexterous manipulation), science operations, and extreme delivery in a Mars‑analog desert environment.
URC at a glance
The University Rover Challenge (URC) is an annual international robotics competition that tasks student teams with designing and operating a Mars rover prototype. Typical missions include autonomous navigation, equipment servicing (dexterous manipulation), extreme delivery, and science operations. Rovers are scored on performance, systems design, and field readiness.
On our team I led systems engineering and integration across mechanical, electrical, and software subsystems, focusing on ROS 2-based autonomy, manipulator control, perception, and test infrastructure.
My Role
- President, Project Manager, and Lead Engineer guiding multi‑disciplinary teams across mechanical, electrical, and software.
- Led ROS 2 integration for autonomy, perception, and manipulator control; built drivers and interfaces for custom hardware.
- Developed simulation workflows and test plans to validate CV/SLAM, obstacle avoidance, and tele‑operations.
System Highlights
- 6‑DoF manipulator with coordinated control for equipment servicing tasks.
- Perception stack for visual odometry/SLAM and waypoint navigation; mapping and obstacle avoidance.
- Ruggedized electronics: motor control, power distribution, sensing, and telemetry integrated into ROS 2 nodes.
My Team
- Our team was one of the smallest in the world, standing at only 8-10 people on a good day. Up against us were teams of 60-100 members from top universities worldwide.
- Despite our size, we consistently punched above our weight class through careful planning, efficient workflows, and a strong focus on reliability and mission readiness.
Big name teams we beat
- Georgia Tech
- Cornell University
- Yale University
- Texas A&M University
- A team of 6 Japanese Universities all working together
- And many more...
I'd like to add that all of these teams were several times our size and in the competition for many years, yet we beat them on our first time in Utah.
System Highlights
- 6‑DoF manipulator with coordinated control for equipment servicing tasks.
- Perception stack for visual odometry/SLAM and waypoint navigation; mapping and obstacle avoidance.
- Ruggedized electronics: motor control, power distribution, sensing, and telemetry integrated into ROS 2 nodes.
Results
- 2024: 24th worldwide placement.
- 2025: 13th worldwide placement.
Three Rovers, Three Iterations
Across three competition cycles, I helped build and led the development of three distinct rover platforms. Each iteration improved reliability, autonomy, and manipulation capability.
Rover I — Foundational Platform
Focus: baseline mobility, tele‑op reliability, and first‑generation manipulator. Established software architecture and telemetry.
For all of us, this was the biggest project we had ever attempted. It was a steep learning curve but laid the groundwork for future success.
The biggest challenge for this iteration was acquiring funding. What you see below was built in just 10 days with my own personal tools and any parts we could scrounge together. The brains for this one was literally my laptop.
It isn't pretty and doesn't work too well, but we learned a lot and proved ourselves to be able to ask for more funding for the next version.
Rover II — Manipulation and Overall Growth
Focus: ROS 2 autonomy stack, improved perception and SLAM, second‑generation 6‑DoF arm with better actuation and control fidelity.
This rover represented a significant leap in reliability and mission readiness. We overhauled the entire machine across mechanical design, software architecture, and electronics.
Still, funding was our biggest limitation.
This rover had a Jetson Orin NX at its core, providing significant computational power for advanced autonomy and perception tasks.
This rover was our first attempt at tank treads, something that significantly improved its off-road capabilities and made us stand out against the other teams. We were and still are the only team in the world to successfully complete all missions with tank treads without major failure.
Rover III — Reliability and Mission Readiness
Focus: robust power distribution, modular electronics, improved tele‑ops UX, and test infrastructure leading to consistent field performance.
This rover was our most reliable and capable platform yet. We focused heavily on systems engineering, test planning, and integration to ensure mission readiness.
With better funding this time around, we were able to procure higher quality components and tools, significantly improving build quality and reliability. Unfortunately, our limitation was now manpower.
The architecture of this rover achieved modularity and ease of development by having separate SBCs for every part of the system. An NVIDIA Jetson Orin NX handled autonomy, perception, and was a WebRTC signaling server for our video streams. It hosted a custom web interface where we could monitor and control the rover remotely and access critical statistics.
I designed and built a custom base station for remote operation. It features 4 3axis joysticks, a touch screen, a Raspberry Pi 5, a wireless router, and some USB PD trickery going on inside. This setup massively improved reliabily and decreased setup time for the competition.
Rover III Key Features
- Modular electronics architecture with dedicated SBCs for autonomy, manipulator control, and telemetry.
- Each camera each used a dedicated PoE SBC for encoding and streaming.
- 6-DoF manipulator arm with improved actuation inverse kinematics.
- Custom ROS 2 nodes for motor control, sensor integration, and mission task execution.
- Robust power distribution system with overcurrent protection and monitoring.
- WebRTC-based video streaming with low latency for tele‑operations.
- Custom web interface for real-time telemetry, control, and mission monitoring.
- Everything starts with systemd services so the robot is plug and play.
- Custom base station with integrated controls and display for streamlined operation.
- The autonomous navigation algorithms are entirely from scratch.
- And so, so much more...
I can share the writeup my team and I did on all the features upon request. I've chosen to not host it here to protect some of the information from our competitors.
Focus: robust power distribution, modular electronics, improved tele‑ops UX, and test infrastructure leading to consistent field performance.