0
Select Articles

Touching the Right Nerve PUBLIC ACCESS

Fly- and Roach Size Micromechanics Probe Some of the World's Most Successful Motion Control Systems.

[+] Author Notes

Associate Editor

Mechanical Engineering 121(02), 54-56 (Feb 01, 1999) (3 pages) doi:10.1115/1.1999-FEB-4

De Ruyter from the University of Groningen in the Netherlands has been studying nervous system of blowflies. These insects have extremely sophisticated visual and tactile perception that is much more sophisticated than our own. De Ruyter placed a tiny tungsten probe into the fly’s head, near one of its motion-sensitive neurons to understand motion control system. With the necessary second probe that acts as a ground (and allows for a current), a nearby oscilloscope feeds the air with an audible indication of voltage. When de Ruyter runs his hand across the fly’s field of vision, the oscilloscope’s output becomes loud and continuous, like old radio static. Davidowitz illustrated how sophisticated these sensory systems are and, therefore, why they would be studied in order to model sense and control systems. Roaches run—when they can—along a wall, with one antenna touching that wall for positioning information. This system checks, feeds information, and makes the control corrections for positioning 25 times a second.

Spending time in the neurophysics lab at the NEC Research Institute near Princeton, N.J. , can make one just a little giddy. On introduction, scientist Rob de Ruyter points to an eight-inch color video display on his benchtop. The screen is filled to its edges with the magnified head of a blowfly (which is like a large housefly). The fly's antennae, exoskeletal hairs, and mouth twitter at a hysterical rate. That's because it is stuck in wax under the scientist's microscope with an electrode in its head, and you start thinking of 1960s science fiction movies.

De Ruyter, who holds a Ph.D. in physics from the University of Groningen In the Netherlands, has placed a tiny tungsten probe into the fly's head, near one of its motion-sensitive neurons. With the necessary second probe that acts as a ground (and allows for a current), a nearby oscilloscope feeds the air with an audible indication of voltage. When de Ruyter runs his hand across the fly 's field of vision, the oscilloscope's output becomes loud and continuous, like old radio static.

And what's even more intriguing, as de Ruyter swipes his hand in the opposite direction in the field, the voltage output is suppressed. He explains that the cell's sense of movement is unidirectional. The probe, when placed near a different cell, would sense movement in the other direction.

The combination of sophisticated science and experimentation, and the feeling of being at play spread a grin across the faces of visitors and scientists alike. I had a vague recollection of playing after school with those smart, weird kids who used to pull the wings off flies and cage them in cork and toothpick huts.

The roach experimentation setup in Hanan Davidowitz's lab. Two contained and funneled speakers (left) are the source of the air motion differences that the roach picks up. At center is the backless roach with the four-pronged electrode inserted. At right, one can see the water-filled tubes that are part of the hydraulic micromanipulator and the cabling for the data acquisition board that is out of view.

Grahic Jump LocationThe roach experimentation setup in Hanan Davidowitz's lab. Two contained and funneled speakers (left) are the source of the air motion differences that the roach picks up. At center is the backless roach with the four-pronged electrode inserted. At right, one can see the water-filled tubes that are part of the hydraulic micromanipulator and the cabling for the data acquisition board that is out of view.

De Ruyter and his colleague Hanan Davidowitz are biophysicists. They are probing the nervous systems of blowflies-and, in Davidowitz's lab, roaches-for good reason. These insects have extremely sophisticated visual and tactile perception that is much more sophisticated than our own. De Ruyter uses this biology manual description to illustrate the point: "The house fly can stop instantly in midflight, hover, turn itself around its longitudinal body axis, fly with its legs up, loop the loop, turn a somersault, and sit down on the ceiling, all in a fraction of a second."

To be able to do this, de Ruyter notes, flies rely on sophisticated systems to control and stabilize flight, and those systems include very fast, accurate measurement and data compilation.

Roaches have two posterior appendages, called cerci, covered with delicate hairs that act as mechanosensors, Davidowitz explains. The sensors pick up minute changes in the air motion behind them, which can indicate the presence of a predator or food source.

If scientists can gain an understanding of how the electrical signals of this sensory input are processed so efficiently, that understanding can be applied to artificial vision and other areas.

Enter mechanical engineering. De Ruyter and his colleagues at the institute work daily with Kenn Fasanella, the lab's senior engineering associate and senior model builder, to fashion the setup of stages, micro manipulators, and probes (with microscopy) that can find and reach the sources of these minute biological electric signals. The design constraints are not to be underestimated.

First, the insects stay alive. Therefore, when manipulating the probes into the blowfly's brain or the roach's central nerve cord the tolerances are extremely tight, about one-thousandth of an inch per inch of linear motion. In his lab, Davidowitz has to delicately remove the roach's top shell and legs to conduct his experiments. The electrode probe cannot move l110re than a few microns, even when de Ruyter's entire fly and probe setup is being put through the fast, wildly changing mechanization that simulates the insect's flight.

Geoff Lewen, de Ruyter's associate, works extensively on the programs to run a tiny, dc motor-powered flight pattern simulator built at the institute.

Another design constraint is avoiding any heat generation into the live insects. Temperature variations far from 72°F would not be tolerated. The heat could destroy the brain and expand the intra- or extracellular fluid.

The probes themselves must be sophisticated. For extracellular signal pickup, there is the chemically etched tungsten probe, which measures just one or two microns at its tip. To measure impulses inside the cell, researchers use glass microtubes filled with saline solution for conduction. They are set up at the ends of arms that can be manipulated very finely, via hydraulic micro manipulators made by Narishige of Japan or Sutter Instruments of Novato, Calif.

The tungsten probe, about to enter the blowfly's head. A digital camera could not get closer to the setup, but one can make out the two bulges that are the fly's compound eyes, each capturing 5,000 pixels of visual information.

Grahic Jump LocationThe tungsten probe, about to enter the blowfly's head. A digital camera could not get closer to the setup, but one can make out the two bulges that are the fly's compound eyes, each capturing 5,000 pixels of visual information.

Nature even assists the researchers. A 100-by-200- micron hole is cut in the fly's head with a broken razor blade. The tungsten or glass probe is put in place, and a scab actually forms around the electrode and holds it in place. This is important, especially when the fly stage is in flight simulation.

The researchers discovered that they had to be sure the fly was getting fruit rather than the sugar water that can sustain life for a time. The hole would desiccate, or shrivel, unless the fly was nourished with dry fruit that Lewen picks up to feed them.

The researchers discovered that they had to be sure the fly was getting fruit rather than the sugar water that can sustain life for a time. The hole would desiccate, or shrivel, unless the fly was nourished with dry fruit that Lewen picks up to feed them.

In this setup, with the fly's brain probed, it can live for five or six days, enough to gather lots of electrical data. De Ruyter jokes that the fly's sitting in front of the oscilloscope screen overnight munching on fruit or drinking sugar water is somewhat analogous to how we humans sometimes spend our evenings.

The question must be asked: Could the data on the physiological responses be skewed because the fly is stuck in wax under bright lights and a microscope? Wouldn't the normal animal response be panic? The scientists must probe into the fly's motion-sensitive neurons, which work constantly and involuntarily. A recording of the activity to the neurons that signal the fly's wings might indeed record a panic message, de Ruyter said, but that's not so with the vision neurons.

Davidowitz is quantifying and examining the physiological responses from the roach's cerci at its rear, which can track changes in air movements. Four electrodes extend from one probe arm, and they are placed under the roach's main nerve "cable" up and down its back. For this, the roach is stripped of its legs and its back shell is removed, which is oddly comforting to devout roach-haters.

Davidowitz's life is made relatively easier because he breeds, in a standard rubber storage bin, the large Madagascar roaches more common in the American South, and not the tiny German cockroaches that are the bane of many a New Yorker or Chicagoan. (De Ruyter captures his blowflies just outside the Princeton facility and breeds them in a five-gallon bucket covered with netting next to his desk.)

How the Instruments Are Made

Kenn Fasanella, Senior Engineering Associate and senior model maker at the NEC Research Institute, supports approximately 20 researchers. Fasanella has about 1,000 square feet of space on the first floor of the lab, which houses his Intergraph workstation as well as manual and automatic milling machines and lathing equipment. A Fadal CNC 88HS does the CAM work.

Fasanella uses the Ashlar Vellum CAD program Vellum 3D, which he describes as more of a real-time 3-D drafting program than solid modeler. Fasanella pointed to abandoned boxes of AutoCAD 13 and 14 in his offices, and said that he has found using the simpler program meets all of his needs without losing so many man-hours training on the Autodesk AutoCAD programs.

The CAM software working with the Fadal system he uses is Virtual Gibbs. He can set the software to the material he will be milling, and it runs a simulation on his display that takes into account how long that material will take to mill.

Many machine parts are stainless steel, but the fly and roach experiments have been somewhat partial to anodized aluminum so that light reflecting from the metal doesn't interfere with the experiments.

A typical contribution of Fasanella's is the manual stage positioning setup. Colleagues Rob de Ruyter and Hanan Davidowitz build the probes to be manually inserted near or in the nerve cells, which is done with the hydraulic micromanipulators. The vertical post of the stage is held in a ball in socket, in which the ball is a collet. This is simple mechanics, where the ball is slatted and the slats are open on one end. When the ball is not squeezed, its expanse keeps it rigidly in place in the socket. When the ball is squeezed, it becomes smaller and can be moved to a new position.

The cercal hairs n1.easure about 0.1 nUl1 at their base, and each of the cerci has about 200 hairs . Each is a mechanosensor, Davidowitz says, because it translates the motion detected into a nerve "action potential." A nerve cord that becomes exposed when the roach's shell is removed carries all manner of these action potentials from different sensory collectors around the roach's body. Davidowitz concentrates on 14 large nerve cells within that cord.

"Action potential" is a standard neuroscience term describing the current change that occurs when a cell is stimulated and lets positive sodium (Na+) ions into its membrane and then lets positive potassium (K+) ions out of the membrane to restore original polarity. The baseline voltage, amplified for ease of use, is a negative 70 mV, and the potential jumps up above, then equally far below that, before returning to polarity.

Davidowitz has these electrical signals amplified 10,000 times and fed into a Nicolet data acquisition system, a data capture board that is gaining popularity with automotive engineers. On 16 channels, the system records the amplitude and frequency of the electrical signals for a few hours at a time.

But Davidowitz needs to know other critical data to make sense of these measurements. Principally, he needs to know the amplitude of the wind rushes that are provoking these responses. He jokes with de Ruyter that. his experiment is surely superior because this part of it uses a laser.

Davidowitz simulates the cercal hairs with an etched fiber optic hair. With a photodiode plate, he records the movement in the laser beam coming from the optic hair when the wind rushes occur, and so can quantify the air movement and correlate that to the physiological response he is measuring.

Davidowitz illustrated how sophisticated these sensory systems are and, therefore, why they would be studied in order to model sense and control systems. Roaches run-when they can—along a wall, with one antenna touching that wall for positioning information. This system checks, feeds information, and makes the control corrections for positioning 25 times a second. Roaches, he added, make The Guinness Book of World Records for having the fastest moving legs in the animal kingdom.

The blowfly probe on top of a triple-axis flight simulator that Rob de Ruyter, Geoff Lewen, and Kenn Fasanella devised.

Grahic Jump LocationThe blowfly probe on top of a triple-axis flight simulator that Rob de Ruyter, Geoff Lewen, and Kenn Fasanella devised.

Copyright © 1999 by ASME
View article in PDF format.

References

Figures

Tables

Errata

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In