Major breakthroughs in robotic manipulation
22 May 2019
The partners of the Horizon 2020 SOMA Consortium have announced the end of their highly acclaimed project. Launched in 2015 with a mission to advance the state-of-the-art robotic manipulation capabilities for industry, the SOMA project leaves a legacy of major scientific breakthroughs.
The SOMA Consortium includes world-renowned researchers, academics and scientists from the Technische Universität Berlin, Ocado, the University of Pisa, IIT - Istituto Italiano di Tecnologia, Deutsches Zentrum für Luft- und Raumfahrt (DLR, the German Aerospace Center), and the Institute of Science and Technology Austria. Horizon 2020 is the EU’s biggest ever research and innovation programme. SOMA has received funding under Grant Agreement 645599.
The traditional approach to robotic grasping and in-hand manipulation uses rigid hands and considers the object’s environment as an obstacle. Traditional research focused on exploring ways to grasp an object without coming into contact with its surroundings.
The SOMA project marks a paradigm shift in approach by using soft hands that can easily adapt to the shape of the object, and leverage the physical constraints of the environment as opportunities to guide manipulation. This fundamental change in approach was inspired by the ways in which humans use their hands.
“The thousands of trials we performed clearly demonstrated that humans grasp and manipulate objects in very different ways from how roboticists had imagined for 50 years. This research inspired us to use soft, compliant hands and to actively exploit the environment in much the same way that humans do. Today, SOMA hands can perform robust grasping in dynamic, open, and highly variable environments without having to rely on a very accurate perception of the system, or the geometric CAD model of the object,” said Professor Oliver Brock, Head of the Robotics and Biology Laboratory, Technische Universität Berlin.
SOMA hands have soft fingers that are not rigidly connected to one another, enabling the hand to adapt to objects of different shapes and sizes. Compliant hands are also uniquely suited to take full advantage of the object’s environment to facilitate grasping and manipulation, such as allowing resistance from the environment to constrain the object, and resistance from the object to guide the shape of the grasp.
SOMA leaves a legacy of state-of-the art robotic manipulation capabilities that can have vast and far-reaching applications across industry. Consortium partner Ocado is already looking to introduce robotic manipulation capabilities into its operations.
Automating the picking and packing of customer orders is a part of Ocado’s mid-term vision after it has already introduced thousands of bots into its highly automated warehouses. Robotic grasping and in-hand manipulation are critical capabilities towards realising that vision.
“When humans grasp an apple, they do not consciously command each joint of their hand to a certain x,y,z location. Instead, they understand that the shape of the apple requires an all-encompassing grasp and allow the shape of the apple to determine where their fingers come to rest.
SOMA’s compliant hands mimic these principles and use their compliance to exploit the object-environment relationship to their advantage. Furthermore, with SOMA’s soft fingers the risks of imparting damage on delicate items is dramatically reduced, significantly bridging the gap between the research lab and industry. We expect to see robotic hands embodying SOMA principles in production in the not too distant future,” said Graham Deacon, Robotics Research Fellow, Ocado.
The future of robotic manipulation
In addition to leaving a legacy of advanced robotic capabilities, the SOMA project played a pivotal role in bringing together Europe’s leading robotic researchers. This collaboration between expert researchers from both academia and industry has been a fundamental enabler for the huge leap forward in robotic manipulation.