Stairways are a fundamental barrier for autonomous ground vehicles. While wheeled robots excel on flat terrain, navigating multi-step staircases remains one of the most mechanically demanding challenges in mobile robotics. Team 28 set out to address this by designing and building a fully autonomous, six-wheeled stair-climbing robot capable of carrying a payload up the 19-step Engineering Gateway staircase at UC Irvine.
Our design utilizes the rocker-bogie suspension system, which is a solution first developed by NASA for their Mars Sojourner rover to maintain continuous wheel contact across uneven surfaces without the need for active stabilization. The project was driven by a practical need: demonstrating that a compact, low-cost ground vehicle can reliably navigate real-world stair environments, with potential applications in search and rescue, building inspection, and last-mile delivery in infrastructure-limited settings.
After evaluating several existing stair-climbing technologies, including transformable wheels, linkage leg systems, and continuous tank treads, the team selected a six-wheeled rocker-bogie suspension system as the most promising approach. The design was specifically optimized for the Engineering Gateway staircase, with steps measuring 5.75 inches in height and 17 inches in depth, including two extended landings at steps 6 and 13.
The chassis consists of two laser-cut 1/4-inch plywood segments forming the rocker and bogie linkage arms (4-inch front rocker, 8-inch rear rocker, 5.5-inch bogie arm), connected by threaded rods to minimize weight. Wheels were 3D printed in PETG filament at 30% gyroid infill for strength-to-weight efficiency, then dip-coated in Plasti-Dip rubber to maximize traction on concrete stair surfaces.
The drivetrain uses six 12V DC worm-gear motors, one per wheel, selected after torque analysis confirmed they could meet stair-climbing demands within the project budget. The worm gear configuration also provides passive self-locking, preventing the robot from sliding back when pausing mid-climb. After prototyping and testing, the team determined that a simplified direct-wired electrical configuration was sufficient, eliminating the need for motor drivers or control software.
The team performed SolidWorks-based geometric simulation to validate that the chassis could articulate through four distinct climbing phases without contacting the stair edges, with calculated pitch angles of 63.4°, 101.8°, 27.6°, and 49.0° respectively. These were verified during proof-of-concept testing, with measured angles of 60°, 105°, 35°, and 55°, within acceptable tolerance. A three-step proof-of-concept climb was successfully completed during Fall Quarter, validating the geometry before full fabrication.
Development followed an iterative test-and-refine methodology across Winter Quarter, progressing from motor and electronics integration testing through full staircase runs and friction optimization, ultimately culminating in a successful timed climb of all 19 steps.
Team 28 successfully designed, fabricated, and validated a six-wheeled rocker-bogie stair-climbing robot capable of autonomously transporting a standard water bottle up the full 19-step Engineering Gateway staircase at UC Irvine.
Key deliverables produced over the two-quarter project include a fully assembled rocker-bogie vehicle with laser-cut plywood chassis, 3D printed PETG wheels, and a Plasti-Dip rubber coating system; a complete design binder with engineering analysis, SolidWorks CAD models, bill of materials, compliance table, and proof-of-concept test results; and an Annual Design Review poster summarizing the project goals, analysis, fabrication process, and competition results.
During Winter Quarter testing, the team identified wheel traction as the primary performance variable. Multiple payload position configurations were tested to optimize weight distribution, and two full rounds of Plasti-Dip recoating were performed. The second recoat proved to be the decisive improvement, providing sufficient rubber grip on concrete for the robot to pull through all 19 steps consistently. The robot achieved a complete staircase climb in approximately 1 minute and 20 seconds, within the target performance window established at the start of the project.
The team competed in the MAE 151B stair-climbing competition during Finals Week, placing 2nd out of three teams. The project demonstrated that a passive mechanical suspension system, combined with careful friction optimization, can reliably solve a complex terrain navigation problem without the need for active control software or sensor-based autonomy.