Three projects and the truth:
Robotic guitar a division of labor © May 11, 2012, Norwich University Office of Communications
Two engineering students and their professor huddled over a tangle of yellow and green wires in a Norwich University laboratory, puzzled.
Their focus was a guitar simulator from a video game. The plastic, Stratocaster-shaped container was outfitted with fret-board sensors, chip sets and dozens of soldered connections. Matt Corriveau and Sean Scales have spent much of their senior year modifying the simulator to send out a MIDI signal when sensors are touched, indicating the user has fretted a note and plucked a string. Two weeks from the end of the spring 2012 semester, however, the contraption isn’t quite working.
You really have to just get in and try things. You can’t spend too much time researching.
“Do they use 220 ohms with five volts?” murmured electrical and computer engineering [ECE] student Corriveau to himself as he reached for his laptop computer. Prof. Michael Prairie put down the soldering iron, crossed his fingers and touched one of the frets and a wire connection. The result was a satisfying “pop.”
Across the room, ECE major Janice Knoop was also dealing with challenges. Her task was to translate the digital music signals produced by the simulator into electrical pulses. Her computer chips weren’t reading the signal, however, and she’s running out of time to figure out why.
“You really have to just get in and try things,” said Knoop, of Cicero, N.Y. “You can’t spend too much time researching.”
These are the first two pieces of a larger project to build a robotic guitar that can be played by an operator using a remote simulator. John Giannino, a mechanical engineering student, will attempt to convert energy into musical vibration through a mechanical system of solenoids and levers that compress the strings of an actual guitar, producing notes based on the electrical information Knoop is working to produce.
In other words, the notes and fretting performed on the plastic simulator will be reproduced on a real guitar. This is the first stage of a longterm Engineering Department project to create a robotic “band” controlled by remote musicians. Students have discovered there is a lot of work to be done, which can be both frustrating and exciting.
“Whenever we get something working, that motivates us to keep going,” said Scales, an ECE major from Sterling, Va.
Only Giannino, of Derry, N.H., feels confident his end of the project is working, but he expects to log many more hours in the lab. Currently, the metal framework that fits over the guitar’s fret board is a crude version attached to cardboard. Both structure and guitar must to be secured to plywood, and he needs to mill a number of levers and other parts in the machining shop before assembling the prototype.
“I’ve never built anything with lever arms before,” he said. “You really have to take your time.”
Giannino, too, has had setbacks. Part of his job was building the mechanism that would strike strings on the strumming-hand side of the guitar in order to play the fretted notes, but there simply wasn’t time. He will submit a drawing and plans for a future student to use.
Large projects often involve months of research and groundwork with a lot of “ramping up” in the final weeks, according to Prairie. This project is ambitious and brings together such concepts as embedded systems, signal processing and interpretation, and control systems.
“This is a complex problem,” said Prairie. “It’s kind of like a sampling of everything they’ve done.”
Students must learn about dealing with communication, frustration and working as a team, as well as figuring how to break a complex problem into smaller tasks, he said.
Corriveau, who is from Millville, Mass., said this came into play when he and Scales had to write code that would recognize and coordinate signals from 30 separate notes on the fret board, and pass on the information as a decipherable signal. It has been a difficult and time-consuming project. A sensor that works on its own might not function when paired with other systems.
“Last night was a High Five,” he said. “That’s when we got all the sensors to work together.”