MIT researchers have designed a robotic finger with a sharp point and tactile sensors to tackle the challenge of identifying buried objects. In experiments, the finger, called Digger Finger, was able to dig through granular media such as sand and rice, and accurately detected the shapes of the submerged objects it encountered. The researchers say the robot could one day perform a variety of underground tasks, such as locating buried cables or disarming buried bombs.

The research will be presented at the upcoming International Symposium on Experimental Robotics. The study's lead author is Radhen Patel, a postdoctoral fellow at MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL). Co-authors include CSAIL doctoral student Branden Romero, Harvard University doctoral student Nancy Ouyang, and Edward Adelson, the John and Dorothy Wilson Professor of Vision Sciences at CSAIL and the Department of Cognitive and Brain Sciences.

The search for objects buried in granular material—sand, gravel, and other loose particles—is not new. Previously, researchers have used technologies that detect the subsurface from above, such as ground-penetrating radar or ultrasonic vibrations. But these techniques only provide a blurry view of submerged objects. For example, they have difficulty distinguishing rock from bone.

"So the idea is to create a finger that has a good sense of touch and can distinguish between the different things it's feeling," Adelson says. "That would be useful if it were a matter of finding and disarming buried bombs, for example." Bringing this idea to fruition involved overcoming a number of obstacles.

The team's first challenge was a matter of form: The robotic finger had to be thin and have a sharp point.

In previous work, the researchers had used a touch sensor called GelSight. The sensor consisted of a transparent gel coated with a reflective membrane that deformed when objects pressed against it. Behind the membrane were three colors of LEDs and a camera. The lights shone through the gel and onto the membrane, while the camera captured the membrane's reflection pattern. Computer vision algorithms then extracted the 3D shape of the contact area where the soft finger touched the object. The device provided an excellent artificial tactile sensation, but it was uncomfortably bulky.

For the Digger Finger, the researchers reduced the size of their GelSight sensor in two main ways. First, they changed its shape to a thin cylinder with a beveled tip. Then, they removed two-thirds of the LEDs and used a combination of blue LEDs and colored fluorescent paint. "That saved a lot of complexity and space," says Ouyang. "It allowed us to make it so compact." The final product was a device whose touch-sensing membrane measured about 2 square centimeters, similar to a fingertip.

Once the size issue was resolved, the researchers focused on movement, mounting the finger on a robotic arm and digging through fine-grained sand and coarse-grained rice. Granular media tend to clog when numerous particles become trapped in place, hindering penetration. So the team added vibration to the Digger Finger's capabilities and subjected it to a series of tests.

“We wanted to see how mechanical vibrations helped to dig deeper and break through blockages,” says Patel. “We ran the vibratory motor at different operating voltages, which changes the amplitude and frequency of the vibrations.” They found that rapid vibrations helped to “fluidize” the medium, clearing blockages and allowing deeper digging, although this fluidizing effect was more difficult to achieve in sand than in rice.

They also tested various twisting motions on both rice and sand. Sometimes, grains of each medium became stuck between the Digger-Finger's tactile membrane and the buried object it was trying to detect. When this happened with rice, the trapped grains were large enough to completely obscure the object's shape, although the occlusion could usually be cleared with a small robotic movement. Trapped sand was more difficult to remove, although the small size of the grains meant the Digger-Finger could still detect the object's general outlines.

Patel says operators will need to adjust the Digger Finger's movement pattern for different setups "depending on the type of media and the size and shape of the grains." The team plans to continue exploring new movements to optimize the Digger Finger's ability to navigate various media.

Adelson says the Digger Finger is part of a program that expands the areas where robotic touch can be used. Humans use their fingers in complex environments, whether it's to find a key in a pants pocket or to feel a tumor during surgery. "As we improve artificial touch, we want to be able to use it in situations where you're surrounded by all kinds of distracting information," Adelson says. "We want to be able to distinguish between things that are important and things that aren't.".

This research has been funded, in part, by the Toyota Research Institute through the Toyota-CSAIL Joint Research Centre, the Office of Naval Research and the Norwegian Research Council.