Suppose that computers had been invented before the automobile.

Suppose one of the biggest electronics firms in the world had then set out to design a passenger car from the wheels up, using incredibly powerful microprocessors to control steering, braking and engine performance -- approaching the commonplace task of driving as if it were the operation of an advanced jet fighter.

That, it turns out, is exactly what General Motors Corp. is attempting to accomplish with its secret Trilby Project.

And that is why GM's executives are licking their chops at the prospects of sharing technology with their newest acquisition, the Hughes Aircraft Co., a leader in defense research and development.

GM hasn't said much about the Trilby Project, which is described as a five-year, $90 million research effort eventually involving about 100 GM scientists and engineers and headed by 20-year GM veteran Neil Schilke.

But outlines have emerged from recent statements by GM Chairman Roger Smith and other GM executives, who call it an attempt to draw on the escalating power of computers to make a dramatic improvement in the automobile.

There are small, complex computers in automobiles now, enabling engines to run more efficiently with reduced pollution, for example.

The idea behind Trilby, however, is to leap far beyond this level. Instead of adding computers to existing engines or brakes, the Trilby team will be looking for an ideal process for stopping a car, designing new braking methods if necessary, and then equipping cars with new, "smart" microprocessors that help drivers stop cars in emergencies.

Other possible applications could include sensors and computer controls for automatic transmissions to handle gear shifting more precisely. These and other advances could significantly improve automotive mileage and durability.

The Trilby engineers also will investigate ways of equipping the braking system with sensors to link it with other systems in the car, like the engine or steering controls, so that they operate together in the best possible way, said Robert A. Frosch, a former head of the National Aeronautics and Space Administration, who was hired by GM to head its laboratories.

Detroit has not had the computer power even to consider such questions, and GM's recent past includes some embarrassing mishaps in employing existing technology along with its impressive successes.

Now, however, Smith is betting GM's future on technology. "I am a firm believer that the major gains in the automotive industry are going to come in electronics," an exuberant Smith said Wednesday, announcing the Hughes acquisition. "The automobile, in some respects, is still in its infancy . . . . The big thing left is electronics. It's the new frontier."

Smith's embrace of electronics goes to the heart of GM's strategy for regaining the competitive edge in the automotive market that it held until the Japanese onslaught that began in the late 1970s. It explains not only the Hughes acquisition, but also the $2.5 billion GM paid last year for Electronic Data Systems Corp., a leader in the design and management of complex data processing networks for insurance customers and other clients.

EDS is playing a key role in GM's ballyhooed Saturn car project, announced in January. The auto maker's goal is to produce an American-built subcompact car capable of competing on cost, quality and performance grounds with Japanese imports.

An important part of the cost and productivity improvements GM is seeking is supposed to come from a "paperless" factory information system designed by EDS in which all sorts of records, from payrolls and invoices to executive memos, will be written and read on computer terminals.

With the help of the United Auto Workers union, GM is studying a vastly different way of building the Saturn cars by bringing together large, finished parts of the car, or modules, for final assembly.

Meanwhile, a steady procession of technology is being fed into GM plants, including robots, improved machining and metalworking equipment and laser measuring devices to perform quality checks. GM has acquired no less than five small companies that are developing vision systems to enable robots to "see" what they're doing by recognizing specific parts and shapes.

Ultimately, GM intends to construct a unified data-handling network that would link an engineer at a computer-aided design terminal who draws an auto part with the cutting and drilling machines that make it on the factory floor.

But the biggest technological leap for GM appears to be the Trilby Project, with its potential for using computer power to aid the motorist.

"We still have to depend on the driver right now," said Smith, "but that can change."

"Should the steering and brake systems know how fast the car is going and should they somehow respond to that" in an emergency, asked Frosch.

"And if that answer is 'yes,' how do you make sure those systems fit with the responses and desires of the driver?" he continued. He can conceive of computerized braking systems that in an emergency, could monitor a car's speed, the traction of the wheels and whether the car was turning or going straight, and help the driver bring the car to a stop in the safest possible way.

But it also would be necessary to assure that if the systems failed, the driver's ability to stop the car unaided would not be impaired, Frosch said.

The analogy used by GM Vice Chairman Howard H. Kehrl is the design of a rocket or missile.

"If you're going to build a missile, you start out with a mission statement of what the missile is supposed to do" -- how far and fast the missile must travel and the size of its payload, Kehrl said in a recent interview. "Then you try to stretch all the technologies. . . . You lay out all the steps you need to get there and the programs you need to get the problems solved," he said. "And you have an organization that oversees this and makes sure all the systems trade-offs are right."

In the defense jargon, the coordinated control of complex operations such as the firing and guidance of a missile is called "systems engineering." Applying that approach to the separate functions of an automobile is a paramount goal in the Trilby project, said Kehrl.

It is an area in which Hughes' engineers and scientists are skilled. Frosch said no plans have been made to draft Hughes' employes for work on the Trilby Project. "I don't know how we'll go about that. I expect we'll find a way to use their knowledge. One good way to transfer technology is to have people move for awhile."

That may not be a completely painless process. Some of the GM computer employes transferred to EDS after that acquisition rebelled for a time at the abrupt change in corporate culture, and the gap between GM's life style and the more laid-back California version at Hughes is even greater.

Nor is there any way to predict how profitable the sharing of technology will be for GM, which paid premium prices for both EDS and Hughes, according to many Wall Street analysts.

There is no turning back for Smith, however, who is convinced that Saturn, Trilby and GM's other technological forays are essential to the company's long-term prosperity.

Frosch said he expects the benefits from the Trilby project's investigations to emerge bit by bit. "As we begin to understand the pieces of the car , we'll see how to control these subsystems better," he said. Much further off is the linking of systems. "It may be we'll never get there," he said.

One thing is clear. There never will be a GM car named Trilby, said Kehrl.

"Trilby" is the title of a Victorian novel by George du Maurier about a female singer who could not perform except under the spell of a hypnotist. That's a theme of the project, says Frosch -- great performance under powerful control.

GM may be scrapping its iron and steel image for an electronic one, but it is still not about to name a car after a Victorian heroine.