State of play: how cobots can reshape manufacturing

**This story first appeared in the October 2025 edition of AMT Magazine**

Industry 4.0 – the integration of digital technologies into all aspects of industry – has been well established for some time now. But now researchers are looking at ways of using robots with a much older, more unpredictable technology.

The human being.

At the heart of the human-centric approach of Industry 5.0 is collaborative robotics, or “cobotics” – which are designed to work safely alongside human workers, and currently for sale and working at sites across Australia.

Australian Cobotics Centre (ACC) associate director for research training Glenda Caldwell says there’s “a big shift” for researchers to consider. “Industry 4.0 is very much about how industry develops and uses more advanced technologies in their processes,” Caldwell says.

“So the focus of Industry 4.0 is the technology itself, whereas Industry 5.0 is looking at, well, what are the complex problems that we face… and how do we use those technologies in different industry sectors to address those problems?”

“Now we understand we’ve got these technologies, but how do we use them beyond just to produce more stuff? How do we increase human wellbeing, climate, planetary wellbeing?”

CSIRO Robotic Manufacturing Team senior research scientist Peter King says regular industrial robots are almost like power tools – “just it’s so much more efficient and productive to use than hand tools.”

Whereas a cobot is designed with some awareness, whether it’s in sensors, cameras or programming, to be safe to move or work alongside people. At the ACC, Dr Caldwell oversees researchers who investigate how humans and robots might work together. “We spend a lot of our time understanding what the person needs to do… We really want to understand what the person is doing in that task or in that environment, and then working with them to identify the best way to introduce a cobot into their workflow and, you know, getting their input and understanding,” she says.

Advanced manufacturing manager for Weld Australia Cornelius van Niekerk says that the cobots currently on the market are equipped with sensors that can be adjusted tostop before coming into contact with a human.

“So, you don’t need the safety cages, you don’t need all the screens that you have with an industrial robot and all the warning labels as well. But there’s still safety aspects to it – you still need to be aware of the fact that you have a moving machine part near you.”

At Weld Australia, van Niekerk helps build connections between technology suppliers and industry experts through cobot road shows, giving him a front-row seat to observe how industry is responding to the arrival of cobots in the market, and what they’re hoping it might achieve.

“The two big ones our member base typically bring up when I ask them about their biggest challenges are skill shortages and productivity problems,” he says.

“Most of them say to me, ‘Cornelius, if we could double our welding workforce, our number of welders, that would actually solve—we would start running at capacity.'” And the promise of cobots for welding applications is a profound one, he adds.

“Most of the members say to us that they get about one to two hours of actual arc-on time in an eight hour working day, which is the average – but it’s shockingly low,” he says.

This low utilisation occurs because skilled welders spend most of their time on ancillary tasks—joint preparation, materials handling, warehousing and setup work rather than the high-value welding itself. Cobots offer a solution that addresses both productivity and workforce concerns without the fear factor associated with traditional automation.

“A big fear is that these things will replace the welders in the workforce,” van Niekerk acknowledges. “Cobots kind of address that fear because they’re so collaborative, whereas an industrial robot, you put it in a cage with all the safety that goes with it—it has no idea it’s working with a human.

That said, the collaborative nature of cobots means they still require skilled operators. “It still takes a skilled welder to program that robot,” van Niekerk explains. “It can take the repetitive, fatiguing jobs off of a welder’s hands to free up that welder to do more skilled work.” King agrees, saying the cobot can complement the work of an existing welder.

“Ideally what you want to do is, once you’ve done your initial assessment of exactly what needs to be repaired, is to then set the robot to go about doing that mind numbingly difficult sort of work… Sometimes it’s done under conditions where you don’t want to have a human doing that sort of work.”

This human-centric approach extends beyond individual workstations to encompass entire organisational systems.

The ACC research that Caldwell oversees focuses on what she terms “socio-technical systems” that consider human factors, design factors, and extended reality (XR) applications.

“We’re looking at the human factors—how do humans and robots work together, what are the emotional and psychological aspects, how do we design better interfaces and programming methods,” she explains.

As an Australian Research Council supported organisation, Caldwell and the ACC researchers have worked with medical device manufacturers, metal fabricators and construction material producers, while van Niekerk has facilitated implementations across welding and fabrication workshops.

Van Niekerk says cobots right now can be programmed to perform very basic welding – making them suited for straightforward, simple and highly repetitive jobs where they are moved into place. “Some people walk away saying automation is actually not for us at all, because it still takes some time to set the cobot up. You still have to move it into place point by point, so it could take a significant amount of time.”

Australia’s manufacturing sector presents a few challenges for the adoption of cobots and Industry 5.0 more broadly, compared to overseas markets.

Unlike mass production environments, Australian manufacturing is characterised by “mass customisation rather than mass production,” Caldwell says.

“We’re not manufacturing huge assembly lines of cars. Most of our manufacturing is smaller production runs, and that’s where these collaborative robots can provide a lot of benefit.”

Van Niekerk says the fabricators that visit his workshops are looking to secure a return on their investments.

He says the economic case for cobots depends on the application, the duration and complexity of the tasks planned.

“The general rule is, you can rent a cobot, but if you have a job that’s going to run for more than 18 months to two years, then you’re moving into the space where it becomes more economical to buy a cobot,” he says.

But the successful implementation of Industry 5.0 requires a strategic approach, Caldwell says – adding that she aims to show that there will be some value in giving cobots a try.

“They’re not going to solve every problem,” she says.

“I think it’s a slow, incremental introduction—you introduce them to do one specific thing, and once there’s trust and understanding and confidence around what it can do there, then gradually the complexity can increase.”