The Hard Science of Soft Robots

Who doesn’t love Robots? I mean what is not to love. It is a out-of-this-world creation of Artificial Intelligence that completely blows out the human mind. They have many strong suits, but unfortunately, delicacy hasn’t been one of them. The rigid limbs attached to the robot makes it difficult for them to grasp or hold a range of everyday objects without dropping or destroying them.

But recently, some researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have come up with a solution. They found that it might be more convenient to use something more commonly associated with new buildings and Silly Putty - Silicone. At a conference, these researchers demonstrated a 3-D-printed robotic hand which was made out of silicone rubber and was able to lift and handle objects as delicate as an egg. The hand with its three fingers has special sensors to help estimate the size and shape of an object accurately in order to identify it from a set of multiple items.

Robots always have a few limitations in what they can do because it is sometimes hard to interact with objects of different sizes and materials. Since grasping is an important step in being able to do useful tasks, there was an urgent need to develop these soft robots with softer limbs and that could make dynamic grasping possible.

Although soft robots cannot always compete with the hard ones, their silicone limbs can allow them to perform backflips, dance to Bruno Mars' "Uptown Funk," fly, dive, and even walk through volcanoes. Also, they can handle tomatoes without making it a puree and resurface unscathed after being run over by a car, unlike the hard robots.

One downside to the extra flexibility of soft robots is that they often have difficulty to accurately measure the position of the object, or pick it up successfully. This is when the CSAIL team’s “bend sensors” come to the rescue. When the gripper hones in an object, the fingers send back location data based on their curvature. With the help of this data, the soft robot can pick up an object and compare it to the existing clusters of data points that represent past objects. With just three data points from a single grasp, the robot’s algorithms can distinguish between objects as similar in size as a cup and a lemonade bottle. The team essentially wants to develop the same skills in the robot as a blindfolded human who can pick an object, feel it and predict what it is. The team wants to give ‘sight’ to the robot, without them actually being able to see.

And now, a postdoctoral scholar Daniel J. Preston's latest soft invention gives these robots new, complex movements. He has introduced the first soft ring oscillator, which gives soft robots the ability to roll. It is another tool in the toolkit to make these smart, soft robots without any electronics, and without any hard valves. To test the functioning of this soft ring oscillator, Preston and his team created five soft robot prototypes. Each of which uses a single, constant source of air pressure to run three pneumatic actuators (the inverters).

One prototype nudges a ball around a ring. Another undulates a stage to keep two different sized beads rolling against the edge. Eventually, all the smaller beads fall through a hole in the side of the stage and sort themselves out.

The ring oscillator is really helpful for rolling motions. For instance, in order to get the hexagonal foam robot to roll forward, the ring oscillator helps inflate a balloon behind the robot and deflate one in front at the exact same time. The coordinated push-and-release shifts the hexagon forward again and again as the balloons inflate and deflate in perfect sync.

This article does not necessarily reflect the opinions of the editors or management of EconoTimes.