A new way of teaching robots to pick up unfamiliar objects without dropping or breaking them has been developed by researchers at the University of Birmingham.
The method involves training a robotic hand by programming it to pick up an object and then to use information learned in that first grip to grasp and move a whole range of similar objects.
The researchers say this paves the way for robots to be used in more flexible ways and in more complex environments, particularly where humans and robots need to be able to work together.
’Current robot manipulation relies on the robot knowing the exact shape of the object,’ said Jeremy Wyatt, Professor of Robotics and Artificial Intelligence at the University of Birmingham.
’If you put that robot into an unstructured environment, for example if it is trying to pick up an object amongst clutter, or an object for which it doesn’t already have an exact model, it will struggle.
’The programming we have developed allows the robot to assess the object and generate around 1000 different grasp options in about 5 seconds.
’That means the robot is able to make choices in real time about the best grasp for the object it has been told to pick up and it doesn’t need to be continually retrained each time the object changes.’
The researchers taught the robot a specific grasp type, for example, a power grip, using the whole hand to curve around an object, or a pinch grip, which uses two or three fingers. The robot was then able to generalise the grip and adapt it to other objects.
The robotic hands used by the team look very similar to human hands, with five jointed fingers, however, the programming would also work with robots that had other types of hand, such as pincer grips.
The research was carried out within the PACMAN (Probabilistic and Compositional Representations for Object Manipulation) Consortium, funded by the European Union. The consortium is led by Birmingham and also includes the Università di Pisa, in Italy, and the Universität Innsbruck, in Austria.
Alta Innovations, the University of Birmingham’s technology commercialisation office, is currently looking for partners interested in licensing the technology. The University is already working with several companies which are keen to incorporate the technology into their processes.