This section was edited by Associate Editor Jeffrey Winters.

FLAT SPEAKERS, NOT FLAT SOUND

THE BASIC DESIGN OF A LOUDSPEAKER IS AS OLD AS ALEXANDER GRAHAM BELL: alternating electric currents drive a magnet in a coil that’s attached to a membrane. Depending on the design, the coil or magnet moves so the membrane vibrates, creating a sound wave.

While that design has certainly stood the test of time, it is not without its limits. Sound quality, for instance, is closely linked to the size of the membrane. That’s why audiophiles hate using the tiny earbuds included with Apple’s iPods. And though video screens have become as thin as a deck of cards over the past decade, good speakers still require a great deal of bulk.

Engineers in England unveiled a new type of speaker in April that might make all those restrictions obsolete.

A mock-up of the Flat Flexible Loudspeaker, which promises to be as thin as foil and as loud as a jackhammer.

The design is as flat and thin as a sheet of paper. What’s more, the engineers say, it has the ability to project sound across distances without a deterioration of quality or volume.

Called the Flat Flexible Loudspeaker, the device was created at Warwick Audio Technologies, a company aligned with the University of Warwick. It is made up of a laminate of two conducting membranes, one continuous and the other comprising a fine mesh, separated by an insulating layer. As a voltage is applied to the membranes, an electric field forms over the laminate and causes the entire surface to flex. The sound pressure levels from the lightweight speakers, the engineers say, can reach up to 105 decibels—the equivalent of a jackhammer operating less than a meter away.

The speakers have such a thin profile that they can fit in unexpected places, such as beneath the fabric lining the interior roof of a car. They can also be printed on and wrapped around curved surfaces, meaning that the FFL speakers could be placed unobtrusively in public areas such as stores or hotel lobbies.

Another difference between the flat speakers and conventional ones is the way they project sound waves. Sound radiates from conventional speakers, which means that the volume drops the farther one goes from the source. The FFL speakers generate sound across the entire surface in a way that is analogous to a spotlight, so that the sound level remains virtually the same across an auditorium as it is up close.

The company says it is in talks with partners who might be able to mass produce the first products by early next year.

SUPER SNIFFER

One of the most promising methods for detection of trace vapors is via micromechanics. Very small cantilevers oscillate so precisely that the deposition of just a few molecules upon one can change its frequency of vibration.

Unfortunately, while this method reveals the mass quite precisely, that’s just about all it reveals. For engineers trying to build detectors capable of sniffing out traces of explosive chemicals, that isn’t good enough. Some devices using this technique have a hard time discriminating between high explosives and innocuous gasoline residues.

A team of researchers at Oak Ridge National Laboratory in Tennessee and elsewhere recently published findings that suggest a better way to detect potentially explosive chemicals. The key is to heat the suspect molecules while on the cantilever.

The method, which is now being developed into a prototype for field testing, was published in the March 2009 edition of The Review of Scientific Instruments.

The key is the thermal characteristics of certain molecules. The research team, led by Thomas Thundat of Oak Ridge and consisting of scientists from there and the Technical University of Denmark, wanted to make a micromechanical sensor that could be a cheaper alternative to an ion mobility spectrometer, which ionizes minuscule amounts of chemicals and measures how quickly they move through an electric field. While such spectrometers are fast, sensitive, and reliable, they are also cumbersome and expensive.

Thundat and his colleagues realized that they could get the best of both approaches by modifying a standard cantilever-based device so that it could be heated. When air introduced to the chamber passed over the cantilevers, some molecules of vapor stuck to them. The researchers then heated the cantilevers for just 50 milliseconds. While non-explosive material didn’t react over that short span, the explosives tested responded in an easily identifiable manner.

In the published results, Thundat and his colleagues claim to be able to detect less than one billionth of a gram of such explosives as trinitrotoluene, pentaerythritol tetranitrate, and cyclotrimethylenetrinitromine. The team is working to improve and optimize the experimental device, and hopes to begin field tests as early as this fall.

FOLLOW THE SUN

The simplest control system isn’t always the best, but the design for a solar power tracking system that won a recent student contest couldn’t be more elegant. Indeed, the MIT graduate students who won the Making and Designing Materials Engineering Contest worked out a biomimetic system that requires no motors or electrical input to keep photovoltaic cells aligned with the position of the sun in the sky.

The concept, called Heliotrope, is derived from the way sunflowers point their heads toward the sun through the expansion and contraction of cells within their stalks. The changes in the cells are triggered by the warmth of sunlight, and the students—Forrest Liau, Vyom Sharma, and George Whitfield—realized that by utilizing the differential heat created by sun and shade, they could make an artificial “stalk” bend.

After looking at stalks made from such materials as bimetalic strips and various polymers, the team found that the most practical plan was to mount a solar cell on an arch made of aluminum and steel. Because these relatively common metals expand at different rates when heated, a properly designed arch made of them will bend toward a heat source. A tabletop prototype made by the students, for instance, will track a small spotlight held nearby.

For this passive tracking design, the students collected a prize of $10,000.

ECHO LOCATOR

AMERICAN SOLDIERS ON PATROL IN IRAQ AND AFGHANISTAN ALMOST ALWAYS HAVE NUMERICAL AND TECHNICAL SUPERIORITY OVER THE INSURGENTS THEY ARE FIGHTING. But none of that matters if they are confronted with a sniper. One well trained, well placed sniper armed with a simple rifle can pin down an entire patrol.

But technology being developed at the Institute of Software Integrated Systems, part of Vanderbilt University in Nashville, may be able to erase that disadvantage for soldiers. The technology promises to give soldiers the ability to identify not only the three-dimensional location of the sniper, but also the caliber and type of weapon he is shooting.

The system relies on the sound waves produced by a rifle shot. The blast from the muzzle radiates in all directions, while the bullet creates its own sonic boom as it flies faster than the speed of sound. Each weapon makes a slightly different combination of sounds when fired.

A breadboard version of a sniper locating system is shown mounted on a standard Kevlar helmet.

Locator systems now exist that can use the sound waves to find a shooter, but they have limited utility. They are expensive—running as much as $50,000 each—and they often can only track a sniper within a line of sight.

The new technology being developed by Akos Ledeczi, an engineer at ISIS, mounts four small microphones on the helmet of a soldier. When they pick up the sound of a rifle shot, they feed the signal to a handheld processor, which works to calculate the direction the sound came from. The processor uses filtering software to eliminate echoes and other extraneous noises that could throw off the calculation.

If the system detects both the muzzle blast and the shock wave from the bullet, it can provide both the direction and distance of the sniper.

Unfortunately, the precision of the calculations using data from a single set of microphones is not high. But data from two or more sets of microphones can increase accuracy considerably. Using sensors that enable the system to track the relative location of different helmet-mounted microphone arrays, the processor can use the differences in the times the sound reaches each microphone to better narrow down the sound source. According to Ledeczi, if the system can process data from two or more sets of microphones—say, from a few soldiers on the same patrol—it can determine the direction to within a degree and the range to within a couple of  meters, even if the sniper is 300 meters away.

The technology has been tested successfully at the Army’s Aberdeen Test Center. What’s more, it ought to be relatively easy to employ: each microphone set weighs only a few ounces and uses about $1,000 worth of off-the-shelf hardware.

ROBOT FISH

Monitoring the water conditions of a harbor is relatively straightforward: Stick some sensors on a buoy and have a way for those instruments to report their findings back to shore. The problem is that buoys are by their nature immobile. Detect a chemical spill with such a platform and you still have to do some sleuthing to find the source.

Telling tales out of school: This robotic fish is designed to track polluters.

That problem may soon disappear, if the innovative sensor platform developed by BMT Group Ltd., a U.K.-based engineering consultancy, and the University of Essex works as planned. The platform will be able to swim right up to the source of pollution in a body of water and collect data on the culprit.

That’s because the sensor platform is an eerily realistic robotic fish.

These robots swim with the same undulating motion as real fish, and their metal scales make them look like a silvery carp. Unlike some previous attempts at free-swimming robots, these fish-bots have an autonomous navigational system, meaning they will not need a human pilot to guide them around. Instead, each is programmed to sniff out certain chemicals via sensors embedded in its head and return to a central charging hub when its battery pack runs low (approximately every eight hours). At the charging hub, the fish will relay their data wirelessly to a central control center.

The robotic fish also have the ability to communicate with one another via an ultrasonic system.

Huosheng Hu, a robotics professor at the University of Essex, and his students are building five of these fish for a pilot project being funded by the European Commission. Each of the five-foot-long robots costs about $30,000 to construct. The fish-bots will be released into the port of Gijon, in northern Spain, late next year, where they will try to sniff out leaks from ships and pipelines.

If successful, robotic fish could become part of the everyday aquatic population of major ports.

CONTROLS FOR SMARTER MACHINES
By Alan S. Brown

As computing gets cheaper, manufacturing equipment gets smarter. National Instruments Corp. is a company that made its name in analytical instruments built around everyday PCs, provides a case in point.

The company’s CompactRIO is a small, relatively inexpensive controller that has more analog measurement capabilities than there are in a programmable logic controller, according to NI’s senior product manager, Brian MacCleery. The analog sensors and controller work with LabView, NI’s graphical software. Originally designed to create virtual instruments on PCs, LabView now lets engineers model instrumentation and controls, and then use the same model to operate on manufacturing equipment using the company’s CompactRIO controller.

A controller uses nonlinear algorithms to lift trays of wet concrete.

At VAPO Hydraulics in Dadizele, Belgium, Stijn Schacht used LabView and CompactRIO to control four hydraulic cylinders designed to lift 20-ton steel trays of prefabricated concrete onto a 10-shelf storage rack.

It sounds simple, but there are complicating factors: The wet cement sloshes around, and the prefab parts have window and staircase cutouts that make them unbalanced. To prevent spills, the four cylinders must keep the 6-meter-long concrete trays flat within 2 millimeters while ramping up and down to 45 millimeters per second. What’s more, hydraulic systems are nonlinear. This means that, as pistons move and change fluid compression, resonance stiffness and frequency shift as well.

A PLC system would have to work together while continuously adapting control parameters to the position of the cylinders. “This is something that is not possible to realize with PLCs,” Schacht said.

According to MacCleery, “When you have a multiple input/multiple output control system like this, the PLCs tend to fight each other. It’s like two people holding a bowl of milk and walking across the room.”

LabView and CompactRIO create an environment where Schacht could monitor and control everything through a model using nonlinear control algorithms. The system manages the position of the pistons to within 0.1 millimeter, and has protection interlocks to shut down safely in an emergency.

Schacht says it took two months to develop the control system, and VAPO Hydraulics plans to use it in future systems. In the past, MacCleery added, the same level of customization would have required a far more expensive custom circuit board that would have taken much longer to design.