Market demand, investment link to lift coherent research onto fast track

CNY Business Journal (1994-95), Oct 30, 1995 by Hadley, Mark

EAST SYRACUSE--Sometimes in the business world, the course you embark upon is not always the course that you follow. And so it is with East Syracuse-based Coherent Research Incorporated (CRi).

Its executives believe that the eight-year-old company is about to blast into a new orbit. But CRi is not relying on the technology that it originally developed through funding from the Department of Energy and the Small Business Innovation Research program. Instead, the engine of CRi's success is a sophisticated software program that the company first began to work on in the mid-1980s--a program which utilized the then-unique capabilities of a microprocessor-chip design that its founder and CEO, Charles D. Stormon, had created.

What the company developed in 1987 was a chip architecture that greatly increased the capabilities of a computer to recognize patterns using parallel processing.

"Basically, we were a technology in search of an application at the time," declares James Brule, vice president of marketing for the company. "So we went to work and identified 19 possible commercial applications for our technology, and we licensed this technology for producing the chip to Cypress Semiconductor in California."

Stormon reports that the company's principals decided to bank on developing software applications rather than on trying to become a specialized processor manufacturer--a wise decision as it turns out because, within a few years, computer technology had advanced to the point where a specialized chip-architecture was no longer necessary for high-speed pattern recognition.

Key to Success: "SmartMaps"

The key to CRi's current success, and what recently attracted significant investments from St. Paul Venture Capital, of Minnesota, and Software Consolidations, Inc., of Boston, is a software and hardware system that CRi has named "SmartMaps."

The SmartMaps system, Stormon says, really developed from synergy among the fledgling company's primary areas of strength. "We had three strong suits as a company," Stormon notes. "First, we had a parallel-processor architecture that we shipped to eight companies in 1990. Second, we had a lot of knowledge about expert systems (a branch of artificial intelligence) and their development. And, third, we also had a lot of skill and capabilities in neural networks, another branch of artificial intelligence."

SmartMaps really draws on all three of the company's primary capabilities.

"The things we had in 1990 were a hardware technology and a software technology, and we had to decide in which direction we wanted to go," Stormon recalls. He adds that Niagara Mohawk Power Corp. helped CRi answer that question by contracting for a small, research-and-development project to apply expert systems, neural networks, and parallel-processing technology to the problem of converting the information that they kept on paper maps into an electronic database.

Artificial Intelligence

That research led to the creation of what has become SmartMaps, says Stormon. "The idea was to take all the advantages of neural networks and the advantages of expert systems, two complementary sides of artificial intelligence, and meld them together. These branches of artificial intelligence have their own strengths and weaknesses, but together they complement one another."

He explains that expert systems are adept at following complex patterns to find answers, but do so in a very step-by-step fashion, which makes this approach only as strong as the information employed to guide the program.

Neural networks, on the other hand, work more like the human brain. Stormon likens the situation to a human being in that the more times the computer sees a similar problem or condition, the more likely it is to come up with the right solution.

"You program an expert system. You teach a neural network," he points out.

CRi's success led NiMo to become the company's first customer for SmartMaps, in early 1993. SmartMaps was installed in the utility's electric customer-service division at its offices in Liverpool, Buffalo, and Utica.

So, why didn't the company enter its rapid-growth phase then? Because the primary customers for SmartMaps, according to Brule, are utility companies, telecommunications companies, government agencies, and large industrial plants. "When you are marketing this kind of system to those kinds of customers, decisions don't come in days or weeks," he stresses. It generally takes a year to 18 months between CRi's first presentation to a company and the company's decision to sign or not to sign a contract. But now CRi is starting to feel the positive impact of the presentations and all the work that it has put in over the last year and a half.

Stormon reports that the company has grown from about 30 employees in 1994 to 60 now, and that the pace of growth is not likely to slow soon. He and Brule see extensive need for SmartMaps.

Brule avers that the program reduces by 90 percent, the labor involved in converting maps and information on paper into an integrated computer database. And the typical utility company is dealing with about 50,000 maps and hundreds of thousands of related documents and files, and the typical telecommunications firm may be dealing with 500,000 maps and millions of pieces of related information. Obviously, a 90-percent reduction in labor has an enormous impact.