The University Record, May 20, 1998

U-M engineers win Army grant to build artificial eye

By Adam Marcus
College of Engineering

Researchers from the U-M and the University of Texas, Austin, will combine lenses, tiny lasers and tunable light detectors to build a microchip that could help the military conduct the most accurate remote visual sensing yet.

Working under a $1.6 million grant awarded to the College of Engineering by the U.S. Army, Pallab Bhattacharya, director of the Solid State Electronics Laboratory, will lead a team in the design of an artificial eye on a microchip--a first-of-its-kind optoelectronic device capable of sensing and processing light. The device also could have numerous civilian applications in navigation systems for vehicles, robotics and, eventually, people with visual impairment.

The grant, which starts June 1, is part of a three-year multi-university research initiative awarded to the U-M, with the possibility of a two-year extension.

"We are very excited about the opportunity to undertake this important research, through which we hope to understand a little more of how the human eye works," Bhattacharya said.

Bhattacharya's team will take two approaches to building a prototype system. Both will capture light with a variable-focus lens array, and shuttle it to a computer to process the data. The difference between the two versions will be in what happens in between. To preserve the color, the lens system could be followed by a set of microprisms, which will defract the incoming light into three discrete bands--red, blue and yellow. Those beams in turn will pass into an array of photoreceivers that generate electrical signals, which in turn generate light beams at the individual wavelengths through a so-called vertical cavity surface emitting laser (VCSEL). Or, he said, the device could skip the prism step and simply employ specially tuned photoreceivers with a VCSEL that creates the three desired colors. The designs converge again when the lasers strike the processing unit.

Bhattacharya said the artificial eye will improve on the human version in at least two key respects. Whereas nature's design converts light into electronic nerve impulses, then relays those signals to the brain, the U-M proposal relies on lasers to do the bulk of the transmitting. Because light travels far faster than this biological conduction, signals will move more quickly through the man-made device than through an optic nerve. "The scheme will also allow processing of data away from the focal plane, which has several advantages," he said, including fewer problems with overheated circuits and remote capability.

Yet another advantage of the proposed system, he said, is that whereas humans can only see light in the visible portion of the spectrum, the optoelectronic sensor will be able to convert any wavelength of light into usable information. For practical purposes, this means that the device could be used for night vision--in which infrared radiation predominates--and ultraviolet detection, just as well as for full daylight viewing. "The human eye is an engineering marvel, and no one has ever been able to duplicate it," Bhattacharya said. "But these optoelectronic devices will eventually be able to simulate the most vital functions."

While Michigan engineers have already done much of the electronics groundwork for the project, Bhattacharya said there are still several technical problems to solve. These include making variable focal length lenses that, like the eye, can adjust to every depth of field. In addition, the device will have to have a sufficiently wide array of lenses, as well as the proper processing instructions, to achieve peripheral "vision." The detectors will also have to be able to "see" everything from deep shadows to bright sunlight, and sometimes both at once. And finally, the whole system will have to be mounted in such a way that it is not sensitive to movement, either by using minute gyroscopes or similar technology.

"The challenges are many, but I believe we will be able to overcome them and pave the way for other applications with the optoelectronic sensor," he said.

In addition to Bhattacharya, College of Engineering faculty members George Haddad, Clark Nguyen and Sang Lee are involved in the project. The group is collaborating with Prof. Dennis Deppe and his colleagues at the University of Texas, Austin.