The future of robotics—seeing and understanding

The Australian Centre for Robotic Vision (ACRV)—an ARC Centre of Excellence led by Professor Peter Corke and headquartered at Queensland University of Technology (QUT)—received the $80,000USD first prize at the 2017 Amazon Robotics Challenge in Japan with its custom-built robot.

While robotics is about machines that interact with the physical world, computer vision is about analysing and understanding the world through images. Robotic vision expands the capabilities of robots, allowing them to see and understand the world in which they are working.

The research team at ACRV took on the competition, designed to solve a key robotics problem for the commercial enterprise Amazon—the ability to pick up items and stow them in boxes, in an unstructured environment. While Amazon is able to quickly package and ship millions of items through its warehouses each year, the commercial technologies for the use of robotics are still being developed.

ACRV’s innovative solution, ‘Cartman’, can move along three axes, like a gantry crane, with a rotating gripper that allows the robot to pick up items using either suction or a simple two-finger grip. Cartman performed well, with the challenge combining object recognition, grasping and error detection and recovery outperforming 74 other teams during the competition in Japan.

This is just one way the ACRV is applying its technologies in robotics and advanced computer vision to solve real-world challenges for industry, as well as many other applications such as the provision of healthcare, for sustainable food production, and the monitoring of the environment.

	The ARC Centre of Excellence, together with its partners, is leading the world working to develop new robotic vision technologies, creating robots that can see, understand and interact.

 

Image: Australian Centre for Robotic Vision director, QUT Professor Peter Corke with members of the Centre’s Amazon Picking Challenge team and Cartman, their Challenge winning robot.
Image credit: Queensland University of Technology