MIT's X-Hinges software designs 3D-printed lamps and gloves that sense how they are being touched

Researchers at MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) and Tianjin University in China have built design software that turns 3D-printed objects into devices that sense how they are being touched and moved, MIT News reported on 22 October. The tool is called X-Hinges.

Two materials so signals do not mix

Earlier 3D-printed sensing objects used the same conductive material for both the sensors and the wire-like traces inside them, so the two gave off similar signals, like hearing two radio stations at once. X-Hinges automatically embeds two different filaments: one for the sensors, chosen to send the strongest possible signal, and another for the traces, which does not interfere.

Users design an object by choosing and connecting shapes and picking how each part moves. Most earlier methods allowed only one axis of motion; X-Hinges supports up to three, compressing, up and down, and side to side, and places sensors where the object bends, so it records motion in all those directions at once.

“It's kind of like adding a conscious element to everyday objects.” — Jiaji Li, MIT postdoc and CSAIL researcher, senior author

A lamp, a glove and a shark-shaped controller

The team used X-Hinges to make a glove that adjusts to the wearer's hand, records how the finger joints bend and relays that to a humanoid robot hand so it can copy a grasp. They also printed an origami-style lamp that changes brightness when pinched, a surface that recognizes objects placed on it, such as a banana, from touch data, and a game controller shaped like a cartoon shark whose tail works like a joystick.

Lead author Xiang Chang, a Tianjin University PhD student and visiting researcher at CSAIL, said most methods allow only one axis of motion because of signal interference. The paper's co-authors include MIT associate professor Stefanie Mueller.

Calibration is still a hurdle

Each printed copy has a slightly different layout of electrical elements, so two objects that look identical can give different readings. The team built a computer vision system to recalibrate them and hopes to add built-in recalibration. Next, Li wants to combine sensing and movement in a single print. The work will be presented at the ACM Symposium on User Interface Software and Technology in early November.