Uncaged robots: all the rage right now

There’s an apparent explosion of interest in smaller, safer, easy-to-program robots for manufacturers. Brent Balinski takes a look at the fast-growing collaborative robots category.


Factory robots getting smaller, “friendlier”

Acknowledged as the first industrial robot, the Unimate robot arm was installed at General Motors’ now-decommissioned Ewing Township, New Jersey plant in 1961.

The big metal arm has long since been retired – and entered Carnegie Mellon’s Robot Hall of Fame in 2003 – and factory robots have gotten more and more popular since the first of their kind appeared.

Weighing 4,000 pounds (not far off two tonnes) and costing $US 25,000 (about $US 200,000 in today’s terms) the Unimate is also the antithesis of the apparently booming class of collaborative robots. Its step-by-step movements – put to use handling hot die castings and welding tasks – were driven by a primitive magnetic drum memory.

Comparatively elegant, cheap, safe to be around, and able to be programmed by line workers with no coding experience, the new co-bots couldn’t be more unlike their forbears.

They are most obviously “lightweight, easy-to-program and affordable,” Shermine Gotfredsen, General Manager, Asia Pacific (excluding China and India) for Universal Robots, told Manufacturers’ Monthly.

“It’s becoming very affordable for manufacturing – even the SMEs, not just the MNCs,” said Gotfredsen, of some of the almost plug-and-play robot options out there.

Perhaps the two best-known examples of collaborative robots are those made by Gotfredsen’s employer (headquartered in Denmark) and the Baxter Robot, made by the USA’s Rethink Robotics.

Both companies’ products have force sensing technology that mean they will not injure a person who stands in the path of a robot arm.

They generally work on a fixed base (cage-free) alongside humans, handle relatively small payloads, and can be easily programmed through a user-friendly HMI and by manipulating the robotic arm through a set of movements, meaning re-tasking is simple.

Last-month Swiss power and automation giant ABB chose the world’s premier industrial expo, Hannover Messe, to unveil its long-awaited, two-armed YuMi co-bot (which has its origins in the firm's Frida concept robot).

And Japan’s Fanuc has just unveiled its CR-35iA, a comparatively heavy-duty offering.

“A couple of years back you would’ve seen one or two robots in the collaborative robot space; now you find a dozen or more vendors that have robots that they’re certainly calling collaborative robots,” Jim Lawton, CMO of Rethink Robotics, told Manufacturers’ Monthly following the release of his company’s new Sawyer product.

In general, worldwide industrial robot sales in general are booming – up 27 per cent last year, according to preliminary figures for 2014 from the International Federation of Robotics.

However, there’s also no denying the current buzz around this reasonably new category of factory help.

Demand is huge

Universal Robots launched in Australia just last year, citing strong local interest and what Universal could potentially offer smaller manufacturers.

Business has grown steadily since in Australia and explosively in the rest of the region.

Elsewhere in APAC, where Universal set up its regional headquarters in Singapore at the beginning of 2015, industrial robot sales are going through the roof.

The IFR’s figures put Asian robot sales at approximately 140,000 units last year, over 60 per cent of the world’s share. China in particular stood out, with 56,000 purchases (up 54 per cent from 2013).

“This region is a fast-growing region for industrial robots. and this quarter alone for the Asia Pacific region I am operating in, we have already had a 50 per cent growth this quarter compared to last quarter,” explained Gotfredsen.

“So, definitely! [laughs] Definitely it’s a very strong market.”

According to the ambitious Universal, it is on track to reach its goal of doubling revenues every year for 2014 – 2017, and launched a new tabletop product, the UR3, in March at Chicago’s Automate 2015 show.

In the same week UR’s main competitor, Rethink Robotics, launched its own smaller collaborative robot.

The interest in smaller, more nimble, more flexible robots from industries is huge, said Lawton.

Some have said the market might be heading for saturation point, but the amount of unmet demand is enormous, suggests Lawton.

He points to research from Boston Consulting Group suggesting at present there are around nine-tenths of manufacturing tasks that are not yet automated, and a chunk of these can now be tackled.

“This is such a huge and significant space,” he said.

“This is in the US what we call the Wild West. It’s largely uncharted territory.

“And so there will be a bunch of companies – like Rethink Robotics, like Universal Robots, and probably a handful of others – that are very rapidly innovating and bringing new, interesting collaborative robots to market.”

Explaining the boom

Though there is anecdotally a lot of interest in the category, hard, official details on sales aren’t here yet.

“[No,] because this application is just starting,” Gudrun Litzenberger, General Secretary at the IFR, told Manufacturers’ Monthly.

“There are a lot of testing applications in various industries.”

The IFR did however agree that sales would likely increase in the future.

In general, sales are being driven by China, followed by South Korea. The automotive industry, which adopted the first factory robots, is still the most significant sector, followed by electronics.

To highlight, even within Australia, how heavily auto is represented in robot sales, consider that units purchased in Australia were 1,214 in 2012 and seemed to plummet to 323 in 2013.

The main reason in the 2012/2013 difference in volumes was the installation of Toyota’s engine plant in Altona in 2012, the IFR points out in its research.

Co-bot makers argue that their machines make productivity gains available to SMEs those which were previously only available to, say, big automotive companies.

Not every firm can budget and plan for a whole lot of money and time spent on integration if they decide they might want to invest in automation.

“When a manufacturer makes the decision that they want to go automate a task, they may need some robotic arms, some PLCs, a conveyor, wires, string them all together, they need various sensors, part positioning sensors and cameras and 3D sensors,” explained Lawton.

Then comes installation and coding and getting things running reliably. And then there’s error handling.

“That’s often multiple hundreds of hours of programming time that goes into just the error handling component of it,” he said.

The link between automation and productivity is an obvious one, and helps explain why China, with its rising wages, is investing so heavily in importing industrial robots. (It is also investing heavily in developing a robotics industry of its own. Around 16,000 of the robots sold last year in China were locally made.)

The flexibility of robots that can be relocated within a plant and quickly re-programmed for a different task is also suited to smaller companies that manufacture high-mix, low-volume goods.

“New solutions such as interfaces, control units and software allow diverse tasks to be automated even by people without any experience in robotics,” explained Litzenberger.

“This opens up new potential applications for medium-sized companies in diverse industries.”

It has not just been the smaller businesses buying lightweight robots, however. For example, Universal has already sold ten robots to one client alone, Boeing Australia.

Reshoring and robots

“The conversation has certainly changed now, and it’s still early days, but people are seeing that there’s some value there,” Rodney Brooks, CTO, chairman and founder of Rethink, told Manufacturers’ Monthly.

Brooks began Heartland (which later changed its name to Rethink) after noticing the low amount of automation within some parts of China, and simultaneously realised that the seemingly endless pool of labour available to those offshoring wouldn’t be there forever.

He sees the democratisation of industrial robots as something that will shrink supply chains and help bring a lot of manufacturing back closer to where items are consumed.

The adoption of robotics has been linked by many others to a trend of re-shoring production in the US.

“It will be possible to do more manufacturing in more locations of the world where you don’t need that really specialised supply chain,” Brooks said.

Rethink believes that collaborative robots (it generally applies the term “interactive robots” to its Baxter) will be a major disruption in the manufacturing industry.

In a way that intuitive, user-friendly tools (think of the iPhone) have changed the way we work, robots with a little commonsense that can be used by just about anybody might be a tool that does the same thing.

Brooks often cites the PC and the drastic (and overwhelmingly positive) changes it brought to the workplace, with Baxter having potential similarities for those in factories. In the way that a user doesn’t need to be a computer  programmer to use a spreadsheet on a PC, a factory worker doesn’t need to be a programmer to use an interactive robot.

Litzenberger said that the “man-machine partnership” would be important in “advancing intelligent production visions (Industrie 4.0).”

“Here, user-friendly robots are opening up opportunities for automation in differing sectors,” she explained.


With all the possibility and maybe even a little hype around the collaborative robots out there, it’s probably reasonable to remind ourselves that they can’t do everything (and aren’t intended to).

Each model the vendors provide is different, to begin with, in terms of basic things like degrees of freedom, speed, repeatability and payload.

Speaking of payload, this is only a few kilograms for most. (A major point of difference for Fanuc’s CR 35iA – “the world’s strongest collaborative robot – is its ability to lift up to 35 kilograms.)

Different robots have different things they can do well, and things they can do poorly or not at all.

Also, where mobility for a robotic solution is needed, for example, a co-bot is the wrong tool for the job.

“What we have been particularly interested in are robots where you have mobility and you also have manipulation capabilities – in the case of Baxter, for example, it is fixed,” explained Elfes, Senior Principal Research Scientist & Science Leader Robotics, CSIRO, told Manufacturers' Monthly.

“And I think different solutions have different applications."

There are situations where a standard, fixed robotic arm will absolutely make sense and will provide what is needed, and situations in which it will not.

Another limitation for all kinds of robots, and not just collaborative types, is dexterity.

Brooks, who is a former Professor of robotics at Massachusetts Institute of Technology and one of the most celebrated roboticists of all time, concedes this.   

“I’m hopeful that now people will start to explore dexterity more than they have in the last 40 years,” he said.

“And we’ll see if progress is made. But you can’t just order up progress on these things. It may take a few years yet before we have much dexterity.”

And in these early days of collaborative robots, actual applications out there in the field in Australia are limited. There have been ten Universal arms sold to Boeing, but there has only been a total of 50 sold in Australia overall.

There are even fewer Baxters out there making things locally, and the first one in Australia operating in an industrial setting was installed at Haigh’s just this year.

All this may say more about our businesses’ conservatism in adapting newer technologies, though.

And for all the attention given to robots and their usefulness in many applications, they remain un-intelligent. There are many places where they will simply never replace human beings in factories.

“Ultimately we don’t have robots today that are as smart and as versatile or as competent as humans,” offered Elfes.

“And personally I don’t think we will ever have robots that are as competent.”

All images supplied besides slider (from globalspec.com), ABB's YuMi (from http://www.abb.com/) and Fanuc's CR-iA (from http://www.sme.org/)

Leave a Reply