Creating Revolutions, which is based in Miami, makes a spill-proof tabletop pager for the hospitality industry, but the product was plagued by a double-digit failure rate due to human assembly errors, says chief innovation officer Einar Rosenberg. So he brought in a Universal Robots model UR3 collaborative robot (cobot for short) that's designed to work alongside humans, not replace them. He used it for three crucial, very precise tasks, such as applying a spot of sealant inside the device's aluminum housing. The failure rate dropped to less than 1 percent. The cobot "saved us from going under," Rosenberg says.

Cobots are the fastest-growing segment in the automation universe. Two players dominate: Universal Robots, based in Odense, Denmark, and Boston's Rethink Robotics, whose co-founder, MIT's Rodney Brooks, is a co-founder of iRobot, maker of the Roomba vacuum cleaner. Cobots generally don't need complicated coding, and they're cheap (less than $50,000 all in). Or you can rent one. On the downside, cobots can't lift more than a few pounds or reach more than a few feet. They're slow--otherwise they'd outpace their human co-workers--and they're new, so no one knows their actual lifespan, but early adopters are fans. They share their ideas for bringing cobots onto your team.

Pick your spots.

Acorn Sales, a maker of custom rubber stamps in Richmond, Virginia, with a dozen employees and less than $2 million in sales, sells a lot of product online, and its margins are tight. "I'm always looking for ways to get more efficient," says CEO Adam Raidabaugh. Acorn put its $30,000 cobot to work cutting wooden blocks down to size and drilling mounting holes in them--a task it performs better, faster, and cheaper in-house than Acorn's former supplier did. Cobots also excel at picking stuff up over here and putting it down over there (using grippers or suction cups), spot welding, spray-painting, bending, and other such mundane tasks. On the other hand, they're not so good at variable tasks that require discernment or creativity.

Make the proper introductions.

Although 360,000 to 670,000 U.S. jobs have already been eliminated because of robots, according to the Center for Economic and Policy Research, a recent study it conducted found no evidence that "new technologies will make most jobs disappear and humans largely redundant." But try telling that to your staff when a cobot shows up on the shop floor. Rosenberg says he needed four long meetings to convince his employees at Creating Revolutions that people were not going to lose their jobs to cobots. He reminded them that cobots specialize in the dirty, dull, and dangerous jobs. Raidabaugh had a similar conver­sation with his team. "We can now automate tasks no one likes to do," he says. "That frees people up to do more meaningful work," which can create growth opportunities for individuals and the companies they work for. Acorn has never had to lay anybody off in the 53 years since its founding, Raidabaugh says, and it's not about to start now.

 inline image

Assess the risk.

Douglas Peterson, GM of Universal Robots' Americas region, insists he's unaware of a single instance in which a human has been killed, maimed, or otherwise injured by one of his company's machines. Still, he recommends performing a thorough risk analysis before you go live, just as you would with any new piece of factory equipment. Joe McGillivray, CEO of contract manufacturer Dynamic Group in Ramsey, Minnesota, had his doubts. So he programmed his new cobot to take a swing at him. "It hit me in a fleshy part of my body and stopped," he says; no harm done. That said, just because a cobot won't knock you over doesn't mean you can put any old tool in its grip. "An open flame--or a chainsaw--on the end of a robot without any guarding is always unsafe," says McGillivray.

Showcase your cobot.

What Creating Revolutions' Rosenberg was looking for from his cobot was steady, reliable production on the factory floor, which he got. But this cobot, which he named Manuel, turned out to have hidden white-collar skills--sales and marketing. Rosenberg noticed how visiting customers touring his facility invariably gravitated to Manuel, and were transfixed. So Rosenberg made a video of the robot. Now, on sales calls, salespeople first demonstrate the product--a high-tech pager that syncs with customers' cell phones so they can instantly get their server's attention. Then they show the video. Manuel has become "a huge piece of closing sales," Rosenberg says. "When people see the robot in action--every single customer--that puts them over the edge. Their eyes light up." Manuel
is cool, for sure, but Rosenberg also thinks Manuel reassures customers who might be intimidated by his product. "Seeing the robot," he says, "makes them feel like this is being made by top-quality technology."

Finding the right robo-worker.

The best collaborative robots (cobots) have some great qualities, but that doesn't make using them a no-brainer. Consider the following before you buy a mechanical employee.


Rent the future.

Some companies, like Hirebotics of Nashville, rent cobots, thereby turning a cap-ex into an op-ex. "Hirebotics comes down, sets them up, programs everything according to our requirements, and charges on the basis of when the robot is working, just like an employee," says Einar Rosenberg, of Creating Revolutions.

Don't get hooked.

Robots are more precise than humans, but they're not as fast--unless the humans can't take their eyes off them. "Some of our projects took a small dip because the cobots are really fun to watch," says Joe McGillivray, of Dynamic Group. "There's nothing in nature that is observably that consistent."

Do the insurance math.

There is not a lot of actuarial data on cobots--the machines themselves, or how safely they interact with humans. This can make them difficult to insure. Rosenberg was able to find a policy for $800 a year, but many insurers, including State Farm, will not yet write policies on cobots.

Factor in the unknown.

Cobots are inexpensive workers. An entry-level, one-armed model can cost less than $30,000, more for an extra arm, special tooling, or custom coding. But, because they're new, no one can be sure how long these machines will last. The bottom line is that "the cost savings may not be as dramatic as appears at face value," says Jeff Burnstein, president of the Association for Advancing Automation.