Joy Ge

Challenge

Build a robot DJ that can sequence an uploaded playlist without manual input, reach small knobs, faders, and buttons safely, and keep transitions synchronized despite timing error in physical button presses.

Approach

Analyzed tracks with Librosa and NumPy, then used an LLM to select transition timing, controls, and motion sequence. Implemented a closed-loop PID controller to nudge track tempo into alignment and continuously correct synchronization error. Simulated reach and collision constraints in RoboDK, taught control poses, and designed a three-piece EOAT with two TPU contact parts and a 3D-printed holder for knobs and faders.

Impact

Validated autonomous transitions across a 10-song playlist with no manual input. A recorded test held synchronization error to 4 ms, with no audible mismatch in tracks.

Challenge

Redesign an industrial robotic painting cell whose misaligned frame, hardcoded component coordinates, and 45-point calibration made setup slow, difficult to reproduce, and vulnerable to operator error.

Approach

Owned the redesign from CAD and mathematical validation through assembly and commissioning. Rebuilt the breadboard and mounted components around self-locating mechanical features, including repeatable brush mounts. Used SolidWorks Simulation with bolted-hole constraints and a safety factor of 2 to select a thicker board with predicted sag below 1 mm. Developed a partial affine calibration algorithm that uses four probed corners to solve coordinate transforms and two held-out points to verify the result. Integrated the guided workflow, pass/fail feedback, and re-probing into the Python robot controller and operator GUI, and used RoboDK to check reach and collisions.

Impact

Reduced calibration time by 80%, from approximately 50 minutes to 10, while validating less than 2 mm error across representative edge and centre locations on both robot cells. Self-locating mounts also made brush replacement an operator task that could be completed in under two minutes.

Challenge

Build a low-cost robotic arm that follows a wearable controller at a distance while producing stable multi-axis motion from noisy flex-sensor and IMU inputs.

Approach

Designed and printed the arm and gripper, then controlled MG996R servos from a 5 V, 20 A supply using Arduino Bluetooth modules. Mapped flex-sensor threshold crossings to motor actions and used MPU position data to command arm orientation. Added capacitors to filter supply fluctuations and reduced commanded step size for smoother motion. After the original gripper failed because its linkage lacked spacers and the thin tips would not align, redesigned the mechanism for more reliable closure.

Impact

Demonstrated wireless arm control from 10 m away with approximately 30 cm of reach. The project integrated mechanical design, printed hardware, power distribution, embedded sensing, Bluetooth communication, and iterative failure-driven gripper development.