DAM Robotics (Oregon State University)

Oregon State won the University Rover Challenge twice, in 2008 and 2010, then left the competition for about eight seasons while it built one of the strongest records at the Canadian International Rover Challenge. It came back in 2025 and finished 15^th, then 19^th in 2026. The same rover that wins in Canada sits mid-field in Utah, and the missions it loses are the ones with the most precise manipulation. That makes this the cleanest case in the field of a programme built for one task set.

Identity and results

Earlier seasons used a different task set, and their totals are not comparable with the modern four missions. The record the team keeps is:

Between 2018 and 2024 the team appears in no URC finals table S-URC-HISTRESULTS S-URC22-SCORES S-URC23-SCORES S-URC24-SCORES. In the same years it competed at the Canadian International Rover Challenge and won it in 2018, 2019, 2024 and 2025, with third place in 2021 and seventh in 2022 and 2023 S-DAM-ROVERPAGE.

The 2013 row is worth reading beside the modern ones. The rover finished fourth overall with a presentation score of 93 and a terrain score of 78, and scored zero in Equipment Servicing with 10 in astronaut assistance S-URC-HISTRESULTS. Thirteen years later the same mission family produced 10.53 S-URC26-SCORES.

Rover generations

The team uses no generation names at all. The rover is “the Mars Rover” or “the OSURC rover” in every source, and the team’s own page says only that “the rover itself has been iterated on over decades” S-DAM-ROVERPAGE. That is a real limit on what can be said about the lineage: there is no year by which to compare one machine with the next.

What the record does show is a continuity of practice rather than a continuity of vehicle. The codebase exists as one archive per school year from 2017 to 2027, which means there is a dated snapshot of the software for every season even though there is no named chassis S-DAM-GH.

The team sits inside a multi-project club that states 600 students and calls itself the largest student-led club on its campus, with rover, underwater, drone racing, combat robotics, business and VexU projects and a stocked workspace in Merryfield Hall S-DAM-SITE. The rover is one project among six, which is a different organisational model from a single-purpose rover team.

Subsystems

Scope. The team describes itself as designing and building “every aspect of the rover”, from custom circuit boards to a six-axis arm to autonomous navigation software S-DAM-ROVERPAGE. That is the entire published scope statement. No mass, dimension, wheel, suspension, battery, radio, computer or arm figure appears on the site.

Actuation. The 2026 rover uses ODrive controllers for the arm, drivetrain and gimbal motors, and frameless motors of a second make for the gimbal S-DAM-SAR2026. That is a well-supported commercial combination, and pairing it with a Cartesian motion-planning library implies the arm is intended to be commanded in end-effector space rather than joint by joint.

Software. The stack is ROS 2 Jazzy with MoveIt2 for inverse kinematics and path planning, the depth-camera vendor’s software development kit for point clouds, and a streaming framework for camera video S-DAM-GH-ROVER. A forked version of the motion-planning library sits in the organisation, as does a project for object detection inside the Unity environment.

Operator station. A Unity-based ground station is maintained as its own project with a companion package, and a separate project holds a custom ground-station antenna tower S-DAM-GH. Treating the operator side as two engineered products rather than a laptop is unusual at this level and is the practice most worth copying.

Autonomy. Navigation runs on ROS 2, with a driver for an open-source autopilot localisation stack S-DAM-GH S-DAM-GH-ROVER. Autonomy scored 42 in 2025 and 30 in 2026, both mid-field results that put the team ahead of several better-placed rovers.

Delivery. Nothing resembling a cargo mechanism, winch or drone appears in the team’s material, and the club’s drone racing project is not a rover aircraft S-DAM-SITE. Delivery scored 32 in 2025 and 16 in 2026.

Mission performance

2013: fourth with a zero. Sample 65, terrain 78 and presentation 93 carried a fourth place while Equipment Servicing scored zero and astronaut assistance scored 10 S-URC-HISTRESULTS. The servicing mission was already the weak point in the year of their best modern finish.

2025: the return. The strongest modern result: Equipment 47, the best servicing score in the team’s record, with autonomy 42, Delivery 32 and Science 69 S-URC25-SCORES. A balanced, competent rover across every mission.

2026: the review and the servicing went backwards. The review score fell from 91.38 to 75.86, a loss of 15.5 points, and Equipment Servicing from 47 to 10.53, a loss of 36.5, while Science rose to 88 and autonomy held at 30 S-URC26-SCORES. Losing the written review and the servicing mission in the same season, in a team with this much continuity, points at schedule or personnel rather than at design.

The Canadian comparison. In 2024 and 2025 the team won the Canadian event outright while finishing 15^th at URC in 2025 S-DAM-ROVERPAGE S-URC25-SCORES. The two competitions differ in task set and in field, and the URC missions where this rover loses are the ones that demand the most precise manipulation. Competition-specific practice is not interchangeable, and this is the clearest evidence of it in the field.

Lessons for a first build

• Version by season, not by feature. Nine dated archives, one per school year, make the programme legible to whoever inherits it. Naming them by year rather than by branch also makes it obvious which one won S-DAM-GH.
• Engineer the operator side. A ground station with its own release and a separate antenna-tower project are two small pieces of work that remove a large class of field failures.
• Use the supported commercial stack when manipulation is not the research question. ODrive controllers with a standard motion-planning library and depth-camera software is a fast route to a working arm, and it leaves the team’s own effort for the missions S-DAM-SAR2026 S-DAM-GH-ROVER.
• Source work through the curriculum. Two capstone projects, one on rover design and one on a simulator, show the rover partly built through for-credit coursework. That survives student turnover at the cost of aligning project scope to the academic calendar S-DAM-GH.
• Rehearse the weakest mission. Four Canadian titles did not produce a URC servicing score above 50. A team that wins one competition and loses a mission in another should treat the difference as a schedule item.

What remains uncertain

The most striking gap is that a team with this much published software publishes no hardware numbers: no mass, no dimensions, no wheel count or size, no suspension type, no battery, no radio and no computer board. Every physical claim about this rover would have to come from the review films.

The 2008 and 2010 titles come from the team’s own record rather than from an official result table, so they rest on the team’s account. Why the URC programme paused between 2018 and 2024, and what brought it back in 2025, is a team-side decision that no public source explains. Whether the 2026 rover is the machine that won in Canada in 2025, or a rebuild, is also unpublished. The 2026 regression in the review score may be a late submission or a shortened document, and the score sheet cannot distinguish the two.

Team sources

• osurobotics.club, the club site, covering all six projects.
• Rover team page, the awards table, the six-axis arm and the custom boards.
• OSURoboticsClub on GitHub, 30 archives including one per rover season.
• 2026 System Acceptance Review, the primary film for the 2026 rover.
• 2025 System Acceptance Review, the best modern URC season.