SYDNEY, Feb. 15 (Xinhua) -- Scientists at the Queensland University of Technology (QUT) in Australia have been given a 50,000 U.S. dollar helping hand from tech giant Amazon to continue the research into vision-guided robotic grasping and manipulation.
QUT robotics researcher, distinguished professor Peter Corke, along with research fellow Jurgen "Juxi" Leitner, led the team to victory in the 2017 Amazon Robotics Challenge in Japan, helping their work to gain the attention of the company.
"Real-world manipulation remains one of the greatest challenges in robotics," said Corke.
While significant breakthroughs have been made in developing robot "hands" capable of grasping objects, Corke's team moves the focus from simple grasping, towards the realm of more meaningful vision-guided manipulation.
"In other words, we want a robot to be able to seamlessly grasp an object 'with intent' so that it can usefully perform a task in the real world," he said.
"Imagine a robot that can pick up a cup of tea or coffee, then pass it to you!"
To achieve their goals the team has spent hours monitoring human behaviour to discover how people pick up an object when they need to then pass it to somebody else.
They looked at the way people picked up objects including a pen, a screwdriver, a bottle and a toy monkey, and how the other person would receive that object if they were then to perform a task with it.
Passers tended to leave "handles" available to the receiver to take hold of, an intuitive choice which allows them to then use the object comfortably.
"While most people don't think about picking up and moving objects -- something human brains have learned over time through repetition and routine -- for robots, grasping and manipulation is subtle and elusive," he said.
"We strive to be the world leader in the research field of visually-guided robotic manipulation."