Fifty years ago, the first industrial robot arm (called Unimate) made a uncomplicated breakfast toast, coffee and champagne. While this may have seemed like a seamless achievement, every move and placement was coded with particular care.
Even with today’s more clever and adaptive robots, this task remains tough for stiff-armed machines. They typically only operate in structured environments with predefined shapes and locations and typically cannot cope with uncertainty in placement or form.
However, in recent years, roboticists have overcome this problem by making fingers out of gentle, versatile materials such as rubber. This flexibility allows these gentle robots to lift everything from grapes to boxes and empty water bottles, but they still cannot handle immense and hefty objects.
To assist these gentle robots, researchers at MIT and Harvard University have developed a recent gripper that is both gentle and sturdy: a cone-shaped origami structure that collapses on objects like a Venus flytrap to lift objects up to 100 times its weight. This movement allows the gripper to capture a much wider range of objects – such as soup cans, hammers, wine glasses, drones, and even a single broccoli flower.
“One of my dreams is to create a robot that will automatically bag groceries for you,” says MIT professor Daniela Rus, director of MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) and one of the senior authors of a recent paper on the project.
“Previous approaches to the packaging problem could only handle very restricted classes of objects – very featherlight objects or shape-conforming objects such as boxes and cylinders – but with the Magic Ball gripper system we have shown that we can select – and place a variety of objects, from bottles to wine to broccoli, grapes and eggs,” says Rus. “In other words, hefty items and featherlight items. Items that are dainty, sturdy or with regular or arbitrary shapes.
This project is one of several carried out in recent years in which researchers began to think outside the box when designing robots. Ball-shaped grippersfor example, it can handle a wider range of objects than fingers, but still has the problem of restricted angles. Gentle robot fingers typically exploit compressed air, but are not sturdy enough to lift heavier objects.
Meanwhile, the design of the recent gripper takes a completely different form. The cone-shaped, hollow and vacuum-powered device was inspired by “magic origami ball” and can surround the entire object and effectively lift it.
The gripper consists of three parts: an origami-based skeleton, an airtight shell surrounding the structure, and a connector. The team created it using a mechanical rubber mold and special heat-shrinkable plastic that self-assembles at high temperatures.
The skeleton of the magic ball is covered with a rubber balloon or slim sheet of material, similar to the team’s previous research on fluid-powered artificial muscles inspired by origamiwhich consisted of an airtight skin surrounding a foldable skeleton and fluid.
The team used the gripper in conjunction with a standard robot to test its durability on various objects. The gripper could grip and lift objects up to 70 percent of its diameter, allowing it to grab and hold a variety of gentle foods without causing damage. It can also lift bottles weighing over four pounds.
“Companies like Amazon and JD want to be able to lift a wider range of delicate or irregular objects, but they can’t do it with finger and suction cup grippers,” says Shuguang Li, a joint postdoc at CSAIL and Harvard’s School of Engineering and Applied Sciences. John A. Paulson. “Suction cups can’t grab anything that has holes in it – and they’d need something much stronger than a soft-fingered gripper.”
The robot currently works best with cylindrical objects such as bottles and cans, which could one day become an asset on factory production lines. It’s no wonder that the shape of the gripper makes it tough for it to grab something flat, such as a sandwich or a book.
“One of the key features of this approach to manipulator design is its simplicity,” says Robert Wood, co-author and professor at Harvard’s School of Engineering and the Wyss Institute for Biologically Inspired Engineering. “The materials and manufacturing strategies used allow us to quickly prototype new grippers tailored to the needs of the facility or environment.”
In the future, the team hopes to try to solve the angle and orientation problem by adding computer vision, which will allow the gripper to “see” and be able to grab specific parts of objects.
“It’s a very clever device that uses the power of 3D printing, vacuum and soft robotics to approach the problem of gripping in a completely new way,” says Michael Wehner, an assistant professor of robotics at the University of California, Santa Cruz, who was not involved in the project. “I imagine that in the coming years I will see soft robots, gentle and dexterous enough to pick a rose, yet strong enough to safely lift a hospital patient.”
Other co-authors of the paper are MIT students John Stampfli, Helen Xu, Elian Malkin and Harvard Research Experiences for Undergraduates student Evelin Villegas Diaz of St. Mary’s University. The team will present their paper at the International Conference on Robotics and Automation in Montreal, Canada, this May.
This project was supported in part by the Defense Advanced Research Projects Agency, the National Science Foundation, and Harvard’s Wyss Institute.