Robotic drumstick keeps novices on the beat

A machine that controls a novice's drumstick to help them learn how to play could be the first of a string of robotic musical teachers. The device has also been found to cut the time it takes to pick up new rhythms, according to a study.

Music teachers often guide a student's hand to get across complex or subtle movements, says Graham Grindlay, a computer scientist who developed the device while at Massachusetts Institute of Technology in Cambridge, US. "I had the idea of a drum kit that would guide you through the playing," he says.

Grindlay previously experimented with guiding a drumsticks tip using magnets, but found a mechanical system more effective.

This approach resulted in what Grindlay has dubbed the Haptic Guidance System (HAGUS), which has a drumstick attached to a set of motors. The user grasps the stick with their arm held in position by an adjacent brace (see image, right) so their hand is guided by the action of the motors.

Haptic interfaces use motorised joysticks or other devices to give people physical feedback or let them feel virtual objects as if they were real.

Drum machine

A skilled drummer can use HAGUS to "record" a specific set of beats for it to later teach to a beginner.

When the novice uses HAGUS and tries to play the same beats, the device guides the drumstick using its motors. In this way it can teach users to alter the tempo, by speeding up or slowing down, or to hit the drum with varying strength to match the expert's playing.

Grindlay used 32 people who had no experience of drumming to test HAGUS's effect on learning new rhythms.

He found that when using the robotic teacher, subjects learned how hard to hit the drum 18% more accurately than when they tried to mimic a rhythm after just hearing it. There was also a smaller beneficial effect on people's timing.

Two-handed approach

"More difficult, coordinated motions would be especially good for this kind of system," Grindlay says. "Usually the approach is to learn to play with each hand separately, and then combine them." But he hopes that using HAGUS devices on both hands at once could help people learn that coordination faster.

Chris Chafe of the Center for Computer Research in Music and Acoustics at Stanford University, California, US, says the study does a "wonderful job" of showing how haptics can help learner musicians.

It shows that with the system, he adds, "training goes up and I imagine accuracy of skilled players does, too."

A paper on Hagus will be presented at the IEEE Virtual Reality 2008 conference this week.

Robots - Learn more about the robotics revolution in our continually updated special report.

Original here