Screeching, scary headlines along the lines of “Robots are taking our jobs!” mask a real trend that is emerging: collaborative robots, AKA cobots, which augment, rather than remove, human labour. Compact and highly-flexible, cobots are designed to work safely alongside humans, as opposed to behind a barrier or inside a cage. They are among the fastest growing segments in the robotics market and global sales are expected to reach $3.3 billion in just five years, according to one estimate. So we’ll be seeing a lot more of them very soon.
So what are the implications for the humans who have to learn how to work with these cobots?
Don’t be under the impression that your new robot co-worker is likely to resemble you. To remain cost-effective, the first generation of cobots are basic beasts: walking is one of the hardest actions to automate, so your cobot will be legless. It may be faceless too and have only one arm. That hasn’t stopped workers drawing a face on the lone arm — our human tendency to anthropomorphise is a powerful thing, and something we’ll return to later.
Beyond the factory floor
While collaborative robots have already appeared on factory floors— taking on repetitive, heavy and hazardous work — they’re fast finding their way into other workplaces: surgical robots already perform low-risk operations; robot home-health aids are being developed and driverless vehicles are poised to take over in transportation and trucking.
Food production, food preparation and logistics are tipped as some of the fastest growing areas for cobots. DHL is one of the most recent converts. Four new “Sawyer” robot arms, from Boston-based Rethink Robotics, are now working across 19 of DHL’s co-packing and production logistics centres, deployed on the basis of need.
“The flexible nature of Sawyer allows us to quickly respond to changing needs, delivering solutions to meet demands and fill labour gaps,” explains Simon Woodward, director of co-packing and production logistics at the company.
The human-robot team
It’s easy to see why co-bots are attractive to businesses. They are relatively inexpensive (a basic model from Danish manufacturer Universal Robots sells for around £28,000) and flexible – which is especially useful for small and medium-sized businesses that need to toggle between different products and produce small batches. But what about those who have to work with them?
Early research identified that designing a collaborative robot is really the design of a team: a human-machine cooperative system. To meet a human’s needs, a collaborative robot should be understandable, believable, trustworthy and provide the interactional support expected. But factoring in effective interdependence management and understanding the properties that humans seek when establishing trust — and integrating these into the robot’s decision making — that’s no easy task.
In comparison, developing intuitive interfaces and training programs is relatively straightforward. With a standard industrial robot, programming is usually called for. But the most popular cobots — by manufacturers such as Rethink Robotics and Universal Robots — allow users to simply grab a robot arm and “teach” it by moving it through waypoints added on a touch screen. In a decade the need for a touch screen is even likely to disappear. Thus cobots also lower the entry barrier for smaller companies which have limited capital and expertise to program, calibrate and set up traditional industrial robots.
Catering to this market, Universal Robots offer online training via a free-to-use e-learning facility: the Universal Robots Academy. In the UK, training centres such as the Research and Innovation Facility (RIF) at Bristol Robotics Laboratory and, in the US, the Robotics & Advanced Manufacturing Technology Education Collaborative (RAMTEC) offer businesses free support for collaborative and other robots. RAMTEC even offers training for middle school students on collaborative systems.
But there’s a growing consensus of expert opinion that – whether a robot is for domestic or industrial use, whether it’s helping an engineer prototype a flying car or an elderly person get out of bed – the quality of the human-robot interaction is just as important as core function.
And what this boils down to is trust.
Bridging the gap
Surveys of people’s attitudes to robots and automation give us cause to believe that many people distrust robots. A system that doesn’t inspire trust can’t help anyone, which is why roboticists are working overtime (with robots) to come up with a way to measure trust — a critical step towards being able to promote it.
In the meantime, studies suggest that non-verbal cues, body language, expression and the increasing effectiveness of speech interfaces all help to bridge the gap between man and machine and improve communication between robots and humans.
Currently, Rethink Robotics’ Sawyer is the only widely available robot that makes use of an expressive interface: eyes which look to where it is working and which frown to indicate that parts can’t be found. Chris Harbert, the company’s director of sales for manufacturing solutions points to research showing that this is significant for people, it gives comfort. “Emotion is important,” he says.
Crucially, establishing trust and understanding with a robot is not so very different to building those bonds with a human. It follows that getting humans and robots to work together most harmoniously will result from developers incorporating human characteristics in robotics systems.
And there’s a lot to gain: a MIT study has found that humans and robots working together in a team can be around 85 per cent more productive than teams made of either humans or robots alone. Collaborative robots even have the potential to democratise work, giving people of varying dexterity, size, age and ability an equal shot to excel at physically demanding careers.
The future human-robot team is thus a powerful construct. Some are even worried that the ensuing relationship will become too close. But that’s another story.