Smart gloves that give the wearer the physical sensations of touching virtual objects could enable remote training for precision tasks, such as performing surgery.

The complexity of the hand, from its myriad motions to the subtle senses at our fingertips, makes designing affordable yet high-performance gloves for human–machine interactions very challenging.

Now, a team that included researchers from Soochow University in China has built a stretchy smart glove with electric sensors along each finger and across the palm. One set of sensors detects the motion of picking up a virtual object, while another applies pressure to the hand to simulate touching that object.

The team used machine learning to teach the glove to recognize different shaped objects. They also tested it as a virtual computer mouse, to hold and swing a virtual baseball bat, and in virtual surgical training.

This touch technology could complement voice and image recognition to create a fuller virtual reality experience.