At MIT’s Computer Science & Artificial Intelligence Lab (CSAIL), researchers have developed a series of adaptive training tool prototypes for better learning. Such tools could prove beneficial to people who can’t afford coaches or personal training.
One of the prototypes is a shapeshifting basketball hoop that helps train users’ motor skills. A beginner could start out using the basket at a lower height and with a broader hoop diameter. As they make shots more consistently, the hoop will shrink in diameter and go higher up the pole until it reaches the standard size. However, fail to make those shots, and the opposite will happen.
Other prototypes include an armband that helps golfers keep their arms straight, a bicycle with raisable training wheels, high-heeled shoes that adapt to different scenarios, and adaptive life jackets.
The shapeshifting basketball hoop gives users the ability to improve their skills while continuously adapting to change. Experimental test results showed that training with the adaptive basketball hoop led to better performance than traditional basketball hoops.
A piezo pressure sensor attached to the hoop’s backboard can detect when the ball hits the board; simultaneously, a switch sensor detects when it goes through the hoop.
Currently, the system’s shot accuracy algorithm is basic—it gives users half a point if the ball hits the backboard and one point if it goes through the net. However, the team notes that the hoop could advance and adapt accordingly with more sensors and cameras.
Led by MIT Professor Stefanie Mueller, the team will continue to develop adaptive tools for other uses, such as workplace training, rehabilitation, and even surgical practice.
