Entrepreneurs on a roll

"It's an unbelievable rush to have somebody fall in love with your product and actually be willing to give you money for it," says Gary Shelton, a founder of Exlar Corp., a fledgling startup that is seeing years of hard work begin to pay off.

Sales of the multi-million dollar company's unique roller screw actuator products and standalone roller screws have roughly doubled for each of the past five years, and Exlar now counts among its customers such industry notables as Caterpillar, Ford, DuPont, and Eastman Kodak. A key factor behind this success, claim company engineers, is that they have something that no one else has: a true hydraulic cylinder replacement in a completely sealed, compact, electromechanical package. And they have the patents to prove it.

In many ways, the story of Exlar is the story of a typical entrepreneur who has lived and breathed the dream of many design engineers: Take a product from concept to market with your own successful startup company. And, hopefully, retire rich.

Off and running. The tale of Exlar begins back in 1991 with a typical engineering problem: A manufacturer of a cam-driven packaging machine was looking for an actuator to act as a volumetric piston pump to provide faster and more accurate carton fills. Design engineers also wanted the flexibility to vary the fill rate and blend products without having to redesign cams.

Working as consulting engineers at the time, Gary Shelton, Bernie Raidt, and a few colleagues rose to the challenge. Intrigued by the ability of roller screws to transmit large forces with precise positioning (see next page), they began to play around with them. But they kept running up against the constraint of physical size. To achieve the forces required, a large actuator was required. But just like in Manhattan, real estate is a valuable commodity in packaging machines.

Ultimately, all it required was a little twisted thinking on Shelton's part. "Anytime you start something new there is a moment of inspiration," says Shelton. "And with any real innovation you have to forget everything you know without actually ignoring it, if you know what I mean."

In a typical roller screw configuration (see diagram), a motor is mounted on the end of a long, threaded shaft that is supported by bearings. As the shaft rotates, the nut (or follower) translates back and forth. Shelton inverted the design so that a short, threaded shaft moves up and down a long internally threaded nut. A major advantage is that the bearings do not have to move axially. As a result, the nut is able to function as the armature of the motor, making for a more compact assembly.

And as any engineer knows, anytime you can get a mechanism to serve two or more functions you reduce the number of parts. With a traditional screw, the motor, coupling, and bearings all add to the length of the actuator--and require an additional output (rod or tube) to cover the threaded shaft.

With a customer committed to buying prototypes of the design, Shelton was able to raise the venture capital to fund the development effort. "We would go in and explain why it works and how it works, and you could see their eyes begin to glaze over," recalls Shelton. "But although good investors may not completely understand the technological details, they do recognize that it's the unusual ideas that have the greatest potential. Mostly it was convincing them our concept would work, we were committed to making it work, and that there was a market for the product. Obviously having a customer that was pushing for it was a tremendous help."

It turns out getting the financial backing was the easy part.

Design challenges. "The roller screw design is mechanically simple, but it is analytically complex because it's a sophisticated 3D problem," says Shelton. "Back then of course we didn't have 3D modeling, so I spent a lot of time trying to figure out how to turn it into a 2D problem so I could get it down on paper."

Shelton recalls spending most of his waking hours--including many sleepless nights--trying to figure out how to build a prototype. The customer was calling up every other day asking when they were going to see a part," says Shelton. "I wouldn't say we ever felt like giving up, but there were days when we were glad when 5:00 rolled came."

The first prototype, which Shelton says functioned as a "pretty good back scratcher," didn't work all that great. The second worked wonderfully. So good, in fact, that the performance specs sounded almost too good to be true. "Compared to a ball screw, we calculated the travel life of our roller screw to be something like 20 times greater. Those are the kind of numbers people look at and say 'I don't believe it,'" says Shelton.

Shelton attributes much of Exlar's early success to long hours, hard work, and, well, dumb luck. "A huge part of any development effort is trial and error, because you have no history to go by, nothing to compare yourself to," he says. "And, obviously, there's a lot of luck involved. I don't think we ever expected to get it right on only the second try."

With that second prototype was born the GS actuator--a robust planetary roller screw mechanism integrated within a dc brushless motor. But shortly after selling its first unit less than a year into the development effort, Exlar was already on to the next challenge: Produce it.

Better, faster, cheaper. Exlar's employees knew that if the product was ever to be a commercial success, they would have to refine the manufacturing process. "The first few prototypes were very expensive to make. We knew we had to come up with a process to make the actuator better, faster, and cheaper," says Shelton.

So he spent the next several months running around town knocking on doors trying to find a machine shop that was willing to fabricate the screw. "You don't just run across an ad that says, 'I'm a creative thread grinder, give me a call.' You have to keep looking until you find a guy who says, 'I've never done that before, but I'm willing to give it a try,'" says Shelton. Ironically, the one company Shelton ultimately found that could do the exquisite machining work required was located a mere mile or so away from Exlar's facility.

Meanwhile, Shelton was having difficulty trying to run the company and still do 90% of the design work. Shortly thereafter, Bill Zerull joined Exlar as its new president. "You can't imagine how weird it is to hire your own boss, but Bill is a super guy and he has the talent to take this company where it needs to go," says Shelton.

"It was obvious to me that this is a great company with a great product," says Zerull, an engineer whose career has focused on small business and product development. "One of its major strengths is the people. It's very much of a team effort--most employees have ownership in the company--and an engineer has an opportunity to get involved in every aspect of the business and quickly see the results of his or her efforts."

Marketing was another challenge, but not quite in the way Exlar had anticipated. Up until 1995, marketing efforts had been made mostly by word of mouth. "When you're a new company you have absolutely no name-brand recognition, sometimes you can't even get your foot in the door. So the trick is to get your part into the design engineer's hands. If you can do that, then they'll buy it," says Shelton.

To broaden its exposure, Exlar made up a small brochure with some product specs and aligned itself with a few sales reps. Unfortunately, they underestimated the market response. "The sales reps started to jump on it right away, but we weren't prepared to handle the business," says Shelton.

The problem was the company did not have a full product line. So engineers began to expand the line to include four frame sizes, with force ratings from 100 to 8,000 lbs, stroke lengths from 3 to 18 inches, and rated speeds from 8 to 40 inch/sec. They also worked to develop configurations that were compatible with more types of brushless motor amplifiers.

Engineers also began working on the FT series, which uses a separately mounted motor and provides speeds up to 60 inch/sec., forces to 40,000 lbs, and up to an 8-ft stroke length.

Full speed ahead. With many new product introductions and since John Walker joined the company in 1996 to head up its sales efforts, Exlar has expanded its sales force nationwide and is off to a good start in Europe.

Zerull's plans for the company include continuing to grow at a high growth rate while expanding the actuator product lines. Plans are also in the works to develop other high performance electro-mechanical products that provide unique and robust solutions to motion control problems.

For Shelton, founding a company has been an eye-opening experience. "You could have the best idea in the world, but you've got to be able to come up with the financing and the market has to be ready. Frankly, though. we think we have the best thing since sliced bread. And we're only beginning to hit our stride."

For more information

To speak with a company representative call 1-800-828-6344, X011 and key in the specific Product Code below:

Linear actuators from Exlar Corp. (P): Product Code 4341

A startup survival guide

- Find a customer first. Few solutions looking for problems actually find any.

- Put together a good team--you can't do it all yourself.

- Don't underestimate the power of enthusiasm in seeking venture capital. It's more important than an actual prototype. Never lose that enthusiasm.

- Demonstrate to investors that the product will work and that a market exists for it.

- The product has to sell itself. Startup companies have no name recognition.

- Synchronize your marketing efforts with product development efforts.

- Use your intuition to tell you when it's time to hire someone else to run the company and fill supporting roles.

- Don't depend on good luck, but expect luck to play a part in your fortunes.

- Count on 5% inspiration and 95% perspiration.

- Create a plan and use it to keep focused--it's too easy to get sidetracked.

How roller screws work

Although roller screws are not a new concept in rotary-to-linear power transmission, the technology is neither widely known nor well understood by most engineers.

As shown in the accompanying diagram, a roller screw consists of rollers arranged in a planetary fashion around a center threaded part generally called a screw. The rollers are either attached in a carriage to the outer piece (the nut) or the inner piece (the screw) such that the rollers move axially with either the nut or the screw.

The threads of the three components are selected such that pure rolling contact between the components is achieved. This rolling contact allows a behavior very similar to that of a ball bearing, which is capable of handling heavy loads at very high speeds. Also, as with ball bearings, there is very little friction inherent in the mechanism, which increases efficiency.

The thread geometries are designed such that there is a very large number of points of contact between the mating parts. This large number produces an even distribution of forces and a low load at each contact point, giving the roller screw very long life compared to other rotary-to-linear conversion devices.