FYI.

This story is over 5 years old.

Tech

The UK Needs to Regulate Its Robots

Pushing for flexible regulation could be a killer asset in cornering the robotics market.
Image: Victor H/Shutterstock

Yesterday, the UK laid out its national robotics strategy with a call to be a world leader in robotic and autonomous systems.

We're sitting on a "robotics goldmine," the Technology Strategy Board told the country. We could be "a world-leading nation in the race to create a new generation of smarter tools and services that we will use in our homes and workplaces."

The report enthuses about the "unique position" the UK is in to claim the robotics market, and with it all the spoils of the first mover advantage. Government strategy in other countries, including the US, China, and Japan, it suggests, has been too sector-specific or technology-focused to reap the full rewards of RAS implementation.

Advertisement

A lot of buzz immediately surrounding the publication of the report focused on the plan to use existing UK "assets" as test beds for robot technology, which is admittedly totally frickin' cool. We're going to send them to decommissioned nuclear plants and down abandoned coal mines! Robots in the wild!

It's an important strategic move, because in order to develop robotics technology to the point where it's actually usable in real life—and especially in places where it could be most useful, such as those unreachable by humans—we'll need to test them out in real-world environments and not just the lab. Defunct old infrastructure, rather than being a hazard or an eyesore, might actually give us an edge.

Sellafield nuclear plant, which was suggested as a potential testing facility for robots. Image: Wikimedia Commons/Chris Eaton

But while that's all very exciting, another point raised by the report is the less sexy issue of regulation. That's especially critical when you're dealing with applications in the sort of environments set out by the testbeds above, where the safety stakes are particularly high. You don't want robots trundling around nuclear reactors willy-nilly.

According to the report, "Regulation and certification will also be a vital part of RAS deployment in many sectors. RAS will operate in hazardous environments and in situations where people are vulnerable and at risk, in care homes, in hospitals, in operating theatres, in mines, under the sea, in the air, in nuclear power stations and in disaster zones. In all of these application areas safe operation is critical."

Advertisement

While it acknowledges the issue as essential, the report is rather vague on how the government intends to develop the regulation of technologies. But really, it needs to be considered ahead of the development of the technologies themselves. What's the use of technology if you're not allowed to use it?

You only have to look at existing tech that's somewhat before its time to see the problems that occur when tardy debate and regulation is forced to play catch-up. Google Glass has faced various bans, particularly in cars; there are still a lot of unanswered legal questions around driverless vehicles; and regulations around hobbyist and commercial drones (particularly in the US, it has to be said) are a bit of a shambles.

"It's very hard for the law to keep the pace of technological development, but the approach should not be that the law waits for the technology to arrive," Andrea Bertolini, a law researcher at SSSA and part of the European-wide RoboLaw project, told me.

Regulating early on doesn't necessarily hold technology back; on the contrary, it can set it up for success. "Some of the issues that the law may identify as problems brought about by new technology can be overcome if you identify them early enough," Bertolini explained.

He gave the example of "privacy by design," an approach encouraged by the UK's Information Commissioner's Office whereby designers integrate compliance with data protection laws into their concepts from the get-go, to avoid problems that tend to occur when they're only considered as an afterthought.

Advertisement

It's very hard for the law to keep the pace of technological development.

But Bertolini says that mindset should ideally extend to all aspects of the law, from potential human rights violations to insurance and liability issues. Take the old autonomous car chestnut: Who's liable if a car with no driver crashes?

And for each different piece of tech developed, new issues arise. I asked Bertolini what exactly constituted a robot in terms of the legal discussion, and he said it was almost impossible to answer.

It conjures images of independent, autonomous systems, but that definition wouldn't cover robotic devices such as bionic prostheses, for instance. And in any case, you'd expect the law to treat a Roomba vacuum cleaner very differently to a Predator drone.

For the purpose of the RoboLaw project, which will suggest regulatory proposals to the European Commission this year, researchers are instead looking at different applications of robotic systems separately. But that takes time, and there's a constant race against the progression of technology itself. If new regulations and guidelines that take into account new technologies aren't developed fast enough, judges have to make decisions based on whatever existing laws they have—which might not be the most appropriate.

Are we ready? I asked Bertolini. Projects like RoboLaw are working on it, he said, but added, "It would probably be quite problematic to let a driverless car vehicle drive on European roads tomorrow morning." Even apparently trivial niggles can cause consternation—like the fact that a vehicle is actually defined as something with a driver.

Advertisement
An autonomous car (in the US). Image: Flickr/Ken Conley

In the strategy report, the UK recognises the opportunity snappy regulation could bring in terms of attracting developers. "Creating the best regulatory environment is key in attracting and creating RAS based products and services to the UK, with safe operations and the right regime to test and de-risk during development," it reads. "Discussions with market regulators and the UK government show that the UK has the will and flexibility to explore these issues in a pragmatic manner."

This willingness to address regulation around robotics proactively may prove to be the UK's killer app—more so than infrastructure assets. If it really can live up to these projections, that is.

The report mentions the need for ongoing discussion with regulatory bodies such as business standards agency BSI and the UK's Civil Aviation Authority (CAA). Richard Taylor, a spokesperson at the CAA, told me there was a dedicated group to discuss regulatory issues around autonomous systems.

Naturally, the CAA's remit in this area is around unmanned aircraft, but Taylor said they could only really move at the pace of the technology itself. "The development of regulations is obviously going to be predicated by the technological developments of the industry," he said.

While there are already small, remotely controlled aircraft (drones) in "non-segregated" airspace, he said the prospect of a full-sized and fully autonomous plane was a way off. And when we get there, Taylor said it probably wouldn't be a quick process.

Advertisement

"Obviously it would be a completely new field, so it is going to require new regulations," he said. Then there's the added complication of integrating EU-certified technology into the regulations of national authorities. "It is quite unchartered water," he added.

Obviously it would be a completely new field, so it is going to require new regulations.

Getting ahead in the regulatory game must be a careful balance between the haste of getting new technologies to market first and the more leisurely pace of addressing legitimate concerns.

Which is all to say that we need to keep the ball rolling. David Lane, chair of the Robotics and Autonomous Systems group that put together the report for the Technology Strategy Board, said his group believed the UK could achieve 10 percent of the global market share of the robotics industry by 2025, but the country isn't at the forefront yet. In fact, robotics expert Noel Sharkey told the BBC we'd "already slipped well behind."

That could, however, be all the more reason to focus first and foremost on regulation. While the UK hasn't really been at the forefront of developing autonomous car tech, for instance (though there is notable work going on such as the RobotCar project at Oxford), it announced a review of regulation and legislation at the end of last year with the aim of staking the country's claim as a testing ground instead. It's a clever way of playing catch-up and attracting companies to the UK's shores.

As the report alludes to, the point at which technologies are taken to market—out of the lab and into the real world—is a pretty sweet spot to corner. "Our vision is to reinforce a RAS ecosystem in the UK that will develop skills and allow ideas and innovation to be created and tested in the market place, ahead of international competitors," it says.

There are a whole host of points that play into the robotics and autonomous systems strategy (not least of all something even more boring than regulation: financing), but if the UK can nail its ambitions to create "a transparent regulatory environment for RAS deployment," it might just carve itself a niche as an incubator, if not an innovator.