4WS Mobile Robot

The problem

A four-wheel-steer robot platform, iterated on by bachelor thesis teams over several years, was unstable and could only be driven with joysticks on its handle — and only on a lab power supply.

What I did

• 01Six of us, split into a hardware and a software group — I was in software.
• 02Rebuilt the wheel modules for a mechanically stable robot, with an FEM-checked chassis, and got it running on battery power.
• 03Wireless driving with a DualShock 4 controller, with four steering modes: front-wheel (2WS) Ackermann, four-wheel (4WS) Ackermann, crab steering and rotation on the spot.
• 04An intent system of lights — brake lights, turn signals and acceleration/deceleration animations — to show people around it what it's about to do.
• 05Groundwork for autonomy: ZED stereo-camera object detection, SLAM-based positioning and a simulated path-follower — working on their own, not yet integrated with the drive control.

Result

A stable robot that passed five of six manoeuvring tests — including parallel parking by crabbing sideways into a pocket and a slalom — and carried 500 N on top of its own 300 N weight with no change in steering or speed.

Tags