Despite not placing in the top three on either the individual pick task or stow task, the team from the Australian Centre for Robotic Vision (ACRV) succeeded in the grand champion combined task to win the overall challenge at the 2017 Amazon Robotics Challenge in Japan.

The ACRV team is made up of researchers, early PhD candidates and undergraduate students who combined computer vision, machine learning and a variety of robotic hardware to successfully complete both pick and stow tasks the fastest.

Team leader Dr. Juxi Leitner said their secret was a Cartesian manipulator that they built from scratch. 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.

“We were the only team with a Cartesian robot at the event. Cartman was definitely a large reason for our success,” Leitner said. “With six degrees of articulation and both a claw and suction gripper, Cartman gives us more flexibility to complete the tasks than an off-the-shelf robot can offer."

Nanyang Technological University of Singapore placed third after winning the pick task, while MIT Princeton of the United States won the stow task.

Amazon said this year’s finalists demonstrated sophisticated solutions combining object recognition, pose recognition, grasp planning, compliant manipulation, motion planning, task planning, task execution, and error detection and recovery to successfully pick and stow unique items. Teams were judged based on how many items were successfully picked and stowed by their robots in a fixed amount of time. A total of $270,000 in prizes were awarded to contestants throughout the four-day competition.

“This year, we made some changes to the Challenge to make it even more difficult and to encourage broader participation from multiple robotics fields—and the response was exciting,” said Joey Durham, contest chairperson and manager of Research and Advanced Development for Amazon Robotics. “The versatility of recognition capabilities in an unstructured environment and the dexterity of grasping mechanisms was truly impressive."