The Robots Are Coming— and They’re Networked
The factory of the future will feature many machines working as one.
Ask any number of manufacturers how soon robots are likely to replace human workers, and you’ll get the same response: The cost of building and maintaining a fully automated facility is exorbitant, wildly risky, and off in the far distant future. But collaborative robotics are proving that the future has arrived, and it isn’t the stuff of dystopian sci-fi. It won’t be long before fully autonomous factories become the norm, creating new opportunities in the marketplace. A host of emerging technologies—including artificial intelligence, smart sensors, sense-and-avoid systems, and articulated robotic joints—are converging in interesting new ways. The result is a connected network of machines that work together, just as we humans do. Rather than a single robot performing all steps of a process, many different machines each excel at just one or a few tasks and then communicate with other robots to start the next segment of work. That’s not unlike the workflow in a labor-intensive factory, but for one key difference: Robots will work around the clock, without a break, in terrible conditions, and without pay.
SoftWear Automation, a startup developed in collaboration with Georgia Tech University, has introduced Sewbots, which can produce a pair of jeans or a T-shirt without any human intervention. This might not sound that impressive—after all, cars can now drive themselves on the highway. But soft textiles present multiple challenges. Fabrics are wildly variable, with thousands of tiny distortions in color, stretch, and weave. Experienced human workers spot anomalies and make adjustments as they work. Doing that with machine-learning algorithms and robots didn’t happen until recently. A Sewbot workline, in tandem with Softwear’s other systems, can produce 1,142 T-shirts—the work of 17 humans—in eight hours.
Sewbo, currently based in Seattle, takes a different approach. Rather than inventing highly specialized robot teams to manufacture garments, Sewbo came up with a process that adds a stiffener to fabrics before sewing, which transforms the material to make it feel more like a thin sheet of hard plastic. Sewbo’s robots sew and finish the stiffened garment, which is then washed and returned to its natural texture.
The plummeting cost of sensors and components, and rapidly accelerating advancement in other areas, will mean lots of other opportunities. Tesla’s Gigafactory will soon employ hundreds of robotic arms and “automated guided vehicles,” essentially mobile robots that transport items from one area to another. Taiwan manufacturing behemoth Foxconn has announced that it will use “Foxbots”—collaborative robots—to perform 30 percent of its electronics manufacturing by 2020. Los Altos, California–based Instrumental builds an optical inspection system that identifies tiny variances during production and can help pull out defective products. Its data-crunching abilities can be used to make the entire manufacturing process more efficient.
Some warn that fully automated factories will obviate many manufacturing jobs, and increase unemployment. But technology has always taken jobs that once only humans could perform—and technology has also forced the creation of new jobs to meet society’s changing demands.
The transition to robot workers is good for business. It’s good business, too. Demand for ever-cheaper products has led companies to offshore manufacturing, which often extracts real costs: In 2013, a rundown Bangladeshi factory that made clothes for the likes of Benetton and Walmart collapsed, killing 1,130 people and injuring 2,500 others. But new robots could move production back to the U.S. without driving up prices. Bringing such factories back home could reduce costs throughout the supply chain, with less money spent on overseas contractors, shipping, and foreign taxes. That means greater profits—and still giving consumers a break on price.