Moving to the Music

Composers usually arrange musical notes to express emotion. To set a mood. To get people dancing. To give life to inspiration. To sell records. A team of scientists at Aalto University in Finland is arranging notes for a totally different purpose—to move objects. Their work isn’t likely to top the charts, but it could bring us closer to game-changing medical technologies like lab-on-a-chip devices and new drug delivery systems. It could also be a means for sorting objects and characterizing materials.

Getting from musical notes to cutting-edge technologies requires a stop in the 1780s. It was during that time that physicist and musician Ernst Chladni developed a technique for visualizing the influence of sound on solid objects. Building on the work of Robert Hooke and others, he sprinkled sand on a plate and then used a bow to “play” the plate like a violin. At certain frequencies, called resonances of the plate, patterns emerged in the sand. These patterns are predictable and illustrate the connection between sound and waves. In fact, techniques based on Chladni’s work are used to shape guitars, violins, and other instruments even today.

A Chladni plate vibrating at 5.289 kHz. Image Credit: Peter Kuiper. Public Domain.

Although the patterns are predictable, how the sand moves the rest of the time—when the surface is vibrating at frequencies that aren’t resonant frequencies—is not very well understood. Most scientists thought this motion was random and couldn’t be controlled. However, new work published today in the journal Nature Communications proves them wrong.

Through several experiments, Aalto researchers Quan Zhou, Veikko Sariola (now at Tampere University of Technology), Kourosh Latifi, and Ville Liimatainen show that the motion of small objects on a vibrating plate can be statistically modelled, predicted, and controlled—even at frequencies that don’t lead to these patterns. In fact, the researchers were able to move multiple objects at the same time, in different directions, and toward different targets with carefully chosen musical arrangements played by just one acoustic source.

The experimental setup centers on a small plate, 50 x 50 x 0.525 mm. The plate sits on top of an acoustical source. Unlike Chladni’s bow, the frequency produced by this source is digitally controlled. First, the team placed several tiny balls on top of the plate. They played a note for 0.5 seconds, which caused the plate to vibrate. The vibrations caused the balls to move. A camera looking down on the plate tracked the position of each ball before and after the note was played. This was repeated for 59 notes in the western musical scale. The notes ranged from C6 (1.047 kHz) up to A#10 (29.83 kHz), which includes the first seven theoretical resonant frequencies of the plate.

After repeating this process 50 times for each note, the team had nearly 400,000 data points that tied particular notes to movements. Although you can’t predict exactly where a ball (or other object) will go when you play a particular note, the combined data give you a pretty good idea. That statistical information is key to controlling the motion of objects on the plate.

Here’s how it works. First, you put the objects on the plate and identify where you want each one to go. Based on their starting positions (determined by the camera) and desired ending positions, a computer program goes through the list of notes and identifies which note is likely to move all of the particles in a good direction. After this note is played, the system cycles through this process again based on the new positions. The process repeats until all of your objects reach their targets. Note that you can direct an object along a particular path by updating its target throughout the process.

The team successfully tested this method on several different kinds of objects, including seeds, candies, electrical components, and even water droplets carried along on tiny transporters. They sorted particles into groups, transformed a diamond pattern into a square, and aligned scattering objects, among other tasks. The number of notes you need to complete a task depends on how complicated it is. More complex tasks require more notes. For example, it took 34 notes to get an arrangement of three objects to their targets, but it took 59 notes for six objects.

In addition to describing their system and experiments, the team’s article goes a long way toward explaining the relationship between the distance an object moves and the length, amplitude, and frequency of the note. The authors also discuss some of the current limitations, challenges, and exciting potential applications of their work. This may not be the song of the summer in your hometown, but its implications are likely to outlast the one that is.

—Kendra Redmond