The Sheriff — Bipedal Robot
// ABSTRACT
The Sheriff is a bipedal robot I designed and built from scratch for a semester-long competition at Duke — a race and dance event requiring the robot to walk autonomously and execute choreographed motions on command. The project spans the full stack: mechanical design in SolidWorks, custom electronics, and Python control software running on a Raspberry Pi. Bipedal locomotion is an inherently unstable control problem, and this project sits at the intersection of mechanical design and real-time control.
SKILLS
TOOLS
Motivation
Bipedal robots are mechanically fascinating and notoriously difficult to control — they're inherently unstable systems that must manage dynamic balance across every motion. The end-of-semester competition at Duke provided a concrete target: build a robot that can walk autonomously and demonstrate controlled motion in a live competition setting.
The Sheriff was designed to meet that challenge from scratch.
Mechanical Design
The robot was designed in SolidWorks with the following priorities:
- Low center of mass: hip and knee actuators positioned to minimize the inertia the control system must compensate
- Structural rigidity: laser-cut and 3D-printed frame members dimensioned to avoid deflection under load
- Serviceability: servo mounts and wiring channels designed for quick access during development and debugging
The leg geometry uses a two-link serial chain (hip and knee joints) per leg, giving two degrees of freedom per limb and allowing forward walking gaits.
Electronics
- Compute: Raspberry Pi (on-board)
- Actuation: RC servo motors at each joint, driven via PWM from the Pi's GPIO
- Power: Regulated 5V supply for compute, separate higher-current supply for servos to avoid voltage droop under load
- Sensing: IMU for orientation feedback (integration in progress)
Software Architecture
The control software is written in Python and structured around a servo abstraction layer that maps joint angles to PWM duty cycles. Above that, a gait sequencer generates the joint angle trajectories for each phase of the walking cycle.
Current implementation status:
- Standing and balance: Functional — the robot maintains upright posture under its own weight
- Rudimentary motions: Working — hip and knee joints can be commanded through a range of angles
- Forward walking gait: In active development — the timing and weight-shift sequence for stable forward locomotion is the current focus
Current Challenges
Bipedal walking requires precise coordination of weight transfer and foot placement. The main engineering challenge is ensuring the robot's center of mass stays within the support polygon at every phase of the gait cycle. This requires both mechanical precision (consistent servo response) and control tuning (gait timing that allows stable weight shifts).
Next Steps
- Tune gait timing for forward walking stability
- Integrate IMU feedback to detect and correct lean
- Implement closed-loop posture correction
- Prepare choreographed motion sequence for the dance competition