Around the world, robots are selling at record levels. According to industry body the International Federation of Robotics, industrial robot sales passed a quarter of a million units for the first time in 2015, following five years of growth at an average of 16 percent per year.

Almost two-thirds of the world’s industrial robots are still found in just two industries – automotive manufacturing and electronics – but according to Tom Bonkenburg, Director of European Operations at independent supply chain engineering consultancy St. Onge Company, a surge in innovation in the sector is poised to dramatically increase their prospects for use elsewhere. That could transform the activities of engineering and manufacturing companies, where low production volumes and highly variable tasks have precluded automation in the past.

“Material handling and robotics has traditionally been a slow-moving sector,” he says. “The automotive industry buys so many robots that suppliers have been fighting each other over a few percentage points of price or reliability. They kept improving the gears and motors a bit, and bringing the price down, but the robot arms that make cars have been broadly the same since the 1980s and 1990s.

“Until recently, robots were blind, they were dumb and they didn’t do anything but repeat the same motions over and over again. Now computer power is finally getting to the point where advanced algorithms, combined with low-cost sensors and better vision systems, mean we can apply robots to tasks that were too difficult to tackle before.”

Often, those tasks are things that are easy for humans, but frustratingly challenging for machines, like identifying an arbitrary object on a shelf or conveyor and picking it up. Such capabilities are coming, says Bonkenburg, citing research efforts like Amazon’s “Picking challenge” competition, and Google’s work

using groups of “cloud robots” that learn how to undertake different tasks before sharing their knowledge with each other.

Tom Bonkenburg

 

Tom Bonkenburg is a partner in St. Onge Company, an international supply chain engineering and consulting firm. For the past 18 years Tom has helped top companies design and implement innovative distribution and manufacturing centers. Many of his consulting efforts have focused on custom automation and robotics within the distribution and warehousing environment.

New workers, new jobs

The ability to adapt to unpredictable working environments will open a host of new robotics applications, like picking and packing operations in distribution centers, says Bonkenburg. But for engineering and manufacturing companies, some of the most exciting developments have more to do with new ways robots can be integrated into the production environment.

“Traditionally, the cost of the robot has been only a small part of the cost of the entire project effort required to support it,” he notes. “You have needed skilled engineers to program the robots, elaborate cages and fences to protect workers, and accurate conveyors to deliver parts.”

“Now companies like Rethink Robotics in the U.S., or Universal Robots in Denmark have developed ‘co-bots’ that can operate safely near human workers without the need for protective cages,” says Bonkenburg. “These robots also take a lot less skill to program, offering iPad-style interfaces and the ability to manipulate the robot to teach it a new task.”

HIGH TECH: The welding robot works on a truss inside the welding cell at Prescient, a Colorado-based company that developed a framing system for commercial construction projects.

Those capabilities make it much easier to redeploy robots to new tasks, or switch them between tasks, says Bonkenburg – a boon for companies that tend to work with small batch sizes and a greater variety of products.

“A small collaborative robot can be powered from a conventional wall socket. It doesn’t need to be bolted to the floor with its own dedicated power supply. You start to see a situation where instead of every robot application being an engineering project, the robot becomes a tool that a worker can use without special training,” he says.

“The price of collaborative robots is coming down to $30,000 or $40,000. It is now possible to buy one and try it in different parts of a facility. People are finding creative uses for these robots that were not originally expected. That’s something that is very different now from the way things were done in the past.”

The continually falling price of robots will reshape the economics of their use in other ways, says Bonkenburg. It may eventually become cost effective for companies to operate “swarms” of robots, with spare units held in reserve to meet periods of high demand, or replace units that fail. That’s especially useful in environments like distribution centers, he notes, where demand peaks are particularly high. It also reduces the need for skilled maintenance staff to keep the robots running. “If you have a fleet of robotic forklifts, and one breaks down, you can push it into a corner and wait for a service engineer, while your other machines carry on with their work.”

I’ve seen much more progress in robotics in the past five years than in the previous ten.
Tom Bonkenburg

On the verge of a revolution

Companies need to proceed with caution, however. Bonkenburg warns there is still a way to go before commercially available robots live up to their emerging potential. “When we look at a lot of real-world applications, we find that the robots aren’t quite there yet,” he says. “The current generation of collaborative robots is still a little too expensive. They are not quite fast enough, not quite accurate enough and they don’t quite do enough.”

The direction of travel, however, is very much the right one. “Every year these robots are getting better and better. I’ve seen much more progress in robotics in the past five years than in the previous ten,” says Bonkenburg. “There are huge sums going into robotics research today, and that’s mainly because affordable computing power has risen to the point where problems that were once considered too hard are now becoming interesting.”

That high rate of progress means engineering and manufacturing companies need to keep a close watch on developments, he says. “Talk to the robotics industry, understand what their products can do and what they can’t. If you have applications that might be appropriate for robotics – maybe because they are repetitive, dangerous, or boring for people – but you rejected the approach before because it was too costly or complex, it might be time to look again.” —  Jonathan Ward

Published: February 2017

Images: PR, Aron Ontiveroz/The Denver Post via Getty Images