Nov 8 2019
Robots can often be seen in today’s warehouses and factories, quickly moving about and transporting tools or items from one station to another.
The new “growing robot” can be programmed to grow, or extend, in different directions, based on the sequence of chain units that are locked and fed out from the “growing tip,” or gearbox. Image Credit: MIT.
In the majority of cases, robots can traverse quite easily across open spaces. However, they find it rather difficult to navigate through constricted spaces to perform various tasks, for example, snaking around the engine parts of a car to unscrew an oil cap, or reaching for a product behind a cluttered shelf.
Engineers at the Massachusetts Institute of Technology (MIT) have now constructed a new robot that extends a chain-like appendage. This appendage is so flexible that it can twist and turn in any required configuration, but it is also sufficiently rigid to apply torque to arrange parts in constricted spaces or support heavy loads.
Upon completing the task, the robot retracts the appendage and extends it again, but it does it at a different shape and length to match the subsequent task.
The MIT engineers designed the appendage based on the way plants grow, during which nutrients are transported in a fluidized form, up to the tip of the plant. There, the nutrients change into solid material to gradually create a supportive stem.
Similarly, a gearbox, or “growing point,” integrated into the robot pulls a loose chain of interlocking blocks into the box. Gears present in this box subsequently lock this loose chain of interlocking blocks together and feed the chain out, unit by unit, as a stiff appendage.
The plant-inspired “growing robot” was recently presented at the IEEE International Conference on Intelligent Robots and Systems (IROS) held in Macau. According to the researchers, cameras, grippers, and other sensors can be placed onto the gearbox of the robot. This would allow it to reach into a shelf and seize a product without affecting the organization of surrounding inventory, or to navigate through the propulsion system of an aircraft and tighten a loose screw, among many different tasks.
Think about changing the oil in your car. After you open the engine roof, you have to be flexible enough to make sharp turns, left and right, to get to the oil filter, and then you have to be strong enough to twist the oil filter cap to remove it.
Harry Asada, Professor, Department of Mechanical Engineering, MIT
“Now we have a robot that can potentially accomplish such tasks,” stated Tongxi Yan, a former graduate student in Asada’s laboratory, who headed the study. “It can grow, retract, and grow again to a different shape, to adapt to its environment.”
The research team also includes Emily Kamienski, an MIT graduate student, and Seiichi Teshigawara, a visiting scholar who presented the study results at the IROS conference.
The Last Foot
The robot’s design is an offshoot of Asada’s work in dealing with the “last one-foot problem”—an engineering term that indicates the last foot, or step, of a robot’s exploratory mission or task. Although a robot can spend most of its time navigating open layouts, the last stage of its mission is likely to involve more agile navigation through tighter and more intricate spaces to finish a task.
Numerous concepts and prototypes have been developed by engineers to deal with the “last one-foot problem.” This also includes robots built from soft and balloon-like materials that grow just like vines to enter through narrow gaps.
However, according to Asada, soft extendable robots like these are not sufficiently strong to support “end effectors,” or add-ons like cameras, grippers, and other sensors that would be required to perform a job, after the robot has navigated its way to its destination.
“Our solution is not actually soft, but a clever use of rigid materials,” added Asada, who is the Ford Foundation Professor of Engineering.
Chain Links
After defining the general functional elements of the growth of plants, the researchers looked for ways to imitate this—in a general sense—in an extendable robot.
The realization of the robot is totally different from a real plant, but it exhibits the same kind of functionality, at a certain abstract level.
Harry Asada, Professor, Department of Mechanical Engineering, MIT
A gearbox was then designed by the team to represent the “growing tip” of the robot. This growing tip was similar to a plant bud, where the tip of the plant feeds out stiffer stem as more nutrients travel to the site.
The researchers then fitted a system of motors and gears inside the box. This system works to pull up a fluidized material—in this example, a bendy series of 3D-printed plastic units interlocked with one another, similar to the chain of a bicycle.
When the chain is fed into the box, it turns around a winch, which, in turn, feeds it via a second set of motors. These motors are programmed to lock specific units in the chain to their adjacent units, and a stiff appendage is eventually created as it is fed out of the box.
By programming the robot, the team can lock specific units together and leave others unlocked, to create specific shapes, or to “grow” in specific directions. During experiments, the researchers successfully programmed the robot to navigate around an obstacle as it grew out or extended from its base.
It can be locked in different places to be curved in different ways, and have a wide range of motions.
Tongxi Yan, Study Lead and Former Graduate Student, Department of Mechanical Engineering, MIT
When the chain is rigid and locked, it is sufficiently strong to support a heavy, one-pound weight. According to the researchers, if a gripper was connected to the robot’s gearbox, or growing tip, the robot can possibly grow sufficiently long to navigate through a constricted space and subsequently apply adequate torque to unscrew a cap or loosen a bolt.
According to Kamienski, auto maintenance is an excellent example of tasks that can be performed by the robot.
“The space under the hood is relatively open, but it’s that last bit where you have to navigate around an engine block or something to get to the oil filter, that a fixed arm wouldn’t be able to navigate around. This robot could do something like that,” concluded Kamienski.
The study was partly funded by NSK Ltd.
Video Credit: MIT.