Robots aren't just for the big guys anymore.
A new breed of so-called collaborative machines designed to work alongside people in close settings is changing the way some of America's smaller manufacturers do their jobs.
The machines, priced as low as $US20,000, provide such companies, small jewelry makers and toy makers among them, with new incentives to automate to increase overall productivity and lower labour costs.
At Panek Precision Inc., a Northbrook, Ill., machine shop, 21 shiny new robots hum as they place metal parts into cutting machines and remove the parts after they are done. It's a tedious and oily task once handled by machine operators, who earn about $US16.50 an hour.
One new robot doubled the output from a machine that was previously operated by a worker "because robots work overnight and don't take lunch breaks and they just keep going," says Gregg Panek, the company's president. In some cases, the robots, which are single articulated arms, can even hold a part while it's getting cut since there is no danger of injury.
Robots have been on factory floors for decades. But they were mostly big machines that cost hundreds of thousands of dollars and had to be caged off to keep them from smashing into humans. Such machines could only do one thing over and over, albeit extremely fast and precisely. As a result, they were neither affordable nor practical for small businesses.
Collaborative robots can be set to do one task one day (such as picking pieces off an assembly line and putting them in a box) and a different task the next.
Some are mobile and able to range freely inside a factory. The use of advanced sensors means they stop or reposition themselves when a person gets in their way, solving a safety issue that long kept robots out of smaller factories.
Small businesses often need flexibility "because they're not just packaging cookies endlessly," says Dan Kara, a robotics expert at ABI research, a market-research firm in Oyster Bay, New York.
At least one company already makes humanoid robots, with arms and computer-screen faces that can flash emotions like surprise. But most lack charisma.
One thing still holding back the trend is fear. Some managers worry that workers will view the machines as competitors for jobs and fight their installation.
Panek says he realised his family-owned machine shop was ready for robots in 2012 after seeing one demonstrated at a machine-tool show in Chicago.
"The guy said you can walk into this thing" and it would stop automatically, he recalls, referring to the sales rep for Denmark's Universal Robots A/S. "Then he let it swivel into him." After the rep wasn't injured, Panek says, "I knew that's what I wanted."
Each robot costs between $US50,000 and $US60,000. Panek plans to add 14 more by the end of next year, he says.
"Having the robots has allowed us to move our existing workers into more useful tasks," such as running more-advanced machines that still require human tending, Panek says. In short, the robots allow one worker to oversee more machines. "We'll always need people."
Instead of operating just one large machine, nine-year Panek employee Jesus Hernandez now supervises several robots.
"At first, I had doubts the robots could do what I did, moving the parts around, handling the parts," says Hernandez, 40 years old. That was until he timed himself against the robots. "I'm fast, but the robots are faster."
Workers at Stuller Inc.'s sprawling jewelry factory in Lafayette, Louisiana, dubbed their newest colleague Fred. He is an autonomous robot that looks like a tall water cooler on wheels that roams a section of the 600,000-square-foot plant, delivering tools to workers at their workbenches and putting the items away when they are no longer needed.
Managers thought the robot might be unpopular, since it eliminated what were once three jobs for human runners.
"But people have taken to him much more readily than we thought," says Jeff High, the company's chief merchandising officer. Some workers actually talk to the machine as if he was human, High says, even though Fred doesn't understand what they are saying or respond.
Mobile robots have been around since the 1950s but they lacked flexibility since they had to follow rigid patterns, usually magnetic lines on the floor or walls. Even now, most systems have to be fenced off so they don't run down wandering humans.
Fred, produced by Adept Technology is programmed with an internal map and sensors that allow it to move through crowded corridors. If its path gets blocked by an unexpected stack of boxes or people on a factory tour, it calculates a new way to reach its destination.
"If you block it, it will find another way, and if you step in front of it again to mess with it, it will say, 'Excuse me, can I get through?'" High says. Stuller plans to deploy a second mobile robot, which costs about $US50,000 for the equipment and related software, in its stonecutting department later this year. They haven't picked a name for it yet.
At K'NEX Brands LLC, a toy maker in Hatfield, Pa., the new robot is called Baxter.
Sold by Rethink Robotics Inc., a Boston company that grew out of research at the Massachusetts Institute of Technology, Baxter has two arms and is designed to roll easily from one spot in a factory to another. It is the most humanlike of collaborative machines now on the market. Baxter (they are all called that) has a computer screen face that can signal emotions. If puzzled, for example, the eyes will cast a quizzical look by arching its eyebrows.
K'NEX, which received the $US20,000 machine in February, has used it for everything from packing boxes to inspecting plastic parts on the assembly line. "When something comes off the line and has to be handled, it becomes labor intensive and that makes us not globally competitive," says Michael Araten, the closely held company's president and chief operating officer. "So it's important to make that automated."
Baxter's current task is putting together window locks, made of three plastic pieces. K'NEX's manufacturing operation makes things other than toys for other customers. Mr. Araten says it is easy to program Baxter, but the process is far from perfect. "Calibrating speed is a key issue," he says, since Baxter's arm has to move at the right speed to catch each part and sometimes misses.
"I look at Baxter like the iPhone one—it was good" but it was just the beginning in the development of smartphones, he says.
This article was first published at the Wall Street Journal. Reproduced with permission.