Adaptive Hoop Gets the Assist

For many people, the best way to learn a new skill is through practice. But when one’s current abilities are a far cry from the ultimate goal, it can be challenging to even get started. In these cases, it can be quite helpful to start with a somewhat easier goal, and work towards the ultimate goal as skills gradually improve over time.

A team at MIT’s CSAIL has been busy building adaptive tools and investigating their impact on motor skill development in their users. They first built an adaptive basketball hoop that can change its height, as well as the diameter of the hoop. The expectation is that a user begins with the basket at a lower height with a larger diameter hoop. As skills improve with practice, the basket can be moved to higher positions, and the diameter of the hoop decreased, until they are in line with regulation measurements.

To be able to continue progressing in skill level, the challenge presented by the device must be neither too easy nor too difficult. The researchers experimented with two different methods for tuning the difficulty level. In the “manually adaptive” mode, users configure the basket as they see fit. The “auto-adaptive” mode, on the other hand, monitors user performance and automatically adjusts the difficulty to an optimal level.

In its prototype form, the auto-adaptive mode is quite simple. A switch sensor is attached to the net to detect when a ball passes through it. In addition to this, a piezo sensor is used to detect the sound of the ball hitting the backboard. Using this sensor information, a scoring system was created in which a scored basket gets one point, a missed shot that hits the backboard gets 0.5 points, and no points are scored for a missed shot. If the user averages at least 0.75 points over four tries, the basket will increase the difficulty level.

The adaptive hoop was evaluated in a small study of twelve participants. The participants in the auto-adaptive study condition were found to have a 25% increase in their average performance score, as compared to the manually adaptive condition. This indicates that people tend to judge their own skills poorly, and often over- or under-challenge themselves.

The team is already hard at work building new devices suited to other applications. In process are devices to assist in learning to ride a bicycle without training wheels, improve golf technique, and walk in high-heeled shoes.