Waikato robocars competing across the ditch

31 August 2017

Autonomous robot

This autonomous robot can reach speeds of 50km/h.

Two teams of robot builders from the University of Waikato are heading to Sydney on Sunday to compete in the National Instruments Autonomous Robotics Competition, going up against 24 teams from other New Zealand and Australian universities.

Team W and Team WESRO, all Waikato engineering students, started working on their projects in January, tasked with designing and building an autonomous robot to manoeuvre around a small-scale ‘city’ mock-up. The city is made up of different areas, including a passenger pick-up zone, a city zone with a speed limit, a passenger drop-off zone and a highway with moving traffic.

Electrical engineer Daniel Dredge is a postgraduate student and team leader for Team W. “I competed in the competition last year so have a fair idea what to expect in Sydney. I’ve been assisting with building the electronics and systems of the vehicle from the ground up, leading the team in design and innovation.”

Daniel has also helped program various systems of the autonomous vehicle, including machine vision, infrared colour detection, ultrasonic distance measurements and vehicle speed control.

“Success in the competition is measured on how fast the robot can complete the course, with as few wall bumps as possible, and your score will also increase the more passengers (pieces of foam) the robot moves from the pick-up and drop-off zones,” he says.

Team W’s vehicle (pictured) is 500mm long and 220mm wide and can reach speeds of 50 km/h. The car is a retrofitted 1/10th remote control car with all sensors necessary to make the vehicle autonomous, which includes a camera, ultrasonic sensors, infrared sensors and a speed controller. “Then for aesthetics, we fitted an RGB battery-level indicator, head, tail and indicator lights to give it the look and feel of a real car,” says Daniel.

The students faced some challenges along the way. “It helped this was my second time, but that didn’t mean it was easy. We decided to use a camera and implement machine vision, which can detect the track, features and objects using filters on an image from a camera. This was challenging because it was something we hadn't done before, so it took us a bit of time to get it right.”

Daniel says the good thing about this competition is the overall challenge to design and build something that is refined enough to compete against other universities. “And the skills you gain designing an autonomous system from the ground up are invaluable in a world that’s becoming more and more automated,” he says.

The winning team in next week’s competition will win two tickets to NIWeek 2018 being held in May in Austin, Texas.