Mechanical Arm Designer | MATLAB Kinematic Model Dev.
Sooner Rover Team (Jan 2026 - Dec 2026)
THE CHALLENGE:
Redesigned the rover arm around carbon fiber components, standardized interfaces, and linear actuators, then developed a MATLAB practice model from the SolidWorks assembly.
During the 2025 University Rover Challenge, the previous arm experienced a wrist-chain failure and an overheated motor. Its mass and maintenance complexity also placed demands on the rover and its operators. The redesign needed to improve reliability and serviceability while retaining the reach and manipulation capability required for competition tasks.
THE SOLUTION:
I developed a modular assembly using custom 3 mm carbon fiber plates and commercially available goBILDA/ServoCity components. A common mounting grid and standardized hardware simplified component interfaces and replacement. Linear actuators controlled the shoulder and elbow, and a smaller linear actuator replaced the chain-driven wrist pitch mechanism. A screw-driven claw provided the end-effector mechanism.
The physical assembly was evaluated using a mock lander panel to investigate motion, access, and mechanical interference. The design documentation records the rationale for component selection, assembly, maintenance, and future improvements so the team can continue development after leadership transitions.
I also developed a MATLAB kinematic prototype using a URDF export and meshes from SolidWorks. The current model animates shoulder and elbow motion and calculates joint angles for requested tip positions. It provides sliders, a side view, and a sampled workspace to support practice and development when the physical rover is unavailable.
RESULTS + NEXT STEPS:
The redesigned hardware was assembled and used in mock lander testing. Full 5 kg payload capability and structural margin remain to be verified; the available documentation does not establish a measured mass reduction or complete structural validation.
The virtual model currently represents two movable joints. Base rotation, wrist and claw animation, actuator stroke constraints, measured joint feedback, and connection to the live rover remain future work. Physical controller access also requires additional setup and testing. Until feedback is integrated, the model should be described as a kinematic practice prototype rather than a validated digital twin.