Robotics could be critical for the UK – if only an Industrial Strategy recognized that

In 2018-20, in the wake of the Brexit referendum, Innovate UK and the Knowledge Transfer Network (KTN) led a series of international Expert Missions to high-tech trade partners, such as the US, South Korea, Japan, Israel, and others, on behalf of the British government. 

One of these was a week-long US Mission by some of the UK’s leading blue-chips, start-ups, and academics in the field of robotics – especially extreme-environment robotics (space, nuclear decommissioning, aerospace, offshore/subsea engineering, and deep mining). I was on that Mission myself, and wrote the subsequent report.

Its aim was to forge alliances and start new conversations with innovators, engineering giants, incubators, and government organizations in the high-tech hotspots of California and Texas – including NASA. The space agency had just announced plans to return to the Moon by 2025, and was also testing the Mars rover and helicopter that are now on the red planet. Delegates were privy to high-level discussions about those projects, and more.

This and other Expert Missions took place under the auspices of the Industrial Strategy and the Industrial Strategy Challenge Fund (ISCF), to help Britain strengthen trade ties and open new doors of opportunity outside the EU. Like the other Missions, the robotics one was also about fact-finding: how do nations like the US both fund and speed innovations out of research labs and into practical applications?

All this and more took place in a heightened, tense, but forward-looking environment. How best could the post-Brexit environment be made to work?

But not long afterwards, disaster struck. The Industrial Strategy – which had only existed since 2017, but was succeeding in its aim of galvanizing investment, linking innovators with funding, and putting robotics and other technologies at the heart of Britain’s plans for the future – was scrapped. 

One reason was the immediate impact of COVID, of course, which the government deemed needed an economic exit plan – the ‘Plan for Growth’. But the other was sheer political vandalism by the Johnson administration, tearing up something that was positive, forward-looking, and – above all – working, but (in his mind) kicked off by his predecessors and political rivals.

There was no rationale behind scrapping a strategy and replacing it with a tactic; both could have co-existed, allowing the Industrial Strategy (and its accompanying investments and networking alliances) to continue, hindered only by pandemic circumstance. 

Yet here we are heading into 2024, and Britain has no coherent, joined-up Industrial Strategy to speak of. And it urgently needs one – updated to reflect current realities, especially in AI. The 2021 Plan for Growth? Well, I’ll leave you to judge whether or not that is working.

From strategy to bullet point 

So, where has all this left the UK’s robotics sector – in which there is world-leading expertise in areas such as space, nuclear decommissioning, subsea engineering, robotic hands and grippers, control systems, and software? Cross-cutting innovations that could be worth billions to the economy, to individual industries, and by boosting jobs and productivity. (A system that solves challenges in one sector is likely to solve similar ones in others.)

Professor James Kell, a Brit, is Robotics Technical Director at engineering multinational Jacobs, which has active interests in both subsea and space robotics. Speaking at a techUK event on autonomous robots this week, he said – in response to diginomica’s question about scrapping the Industrial Strategy:

The ISCF was huge for the robotics community in the UK. Really it started off a big network of people, of university groups across the UK, which still stands via UK-RAS.” [UK Robotics and Autonomous Systems, a division of the EPSRC].

Note that the ISCF also helped give impetus and assistance to the UK’s four academic robotics and AI hubs, ORCA, FAIR-SPACE, RAIN, and the NCNR.

He continued:

The ISCF provided plenty of funding for a lot of different projects, both in UK academia and industry. And it was focused on what we could do that is difficult and different – as opposed to, say, customized automation in assembly lines, like at a car manufacturer. Helping the UK make good on that expertise that we have around the customized, dirty, dangerous places that are difficult.

In short, extreme environments; robotics for where humans fear to tread, or where it is lethal or dangerous for them to do so. For example, in nuclear decommissioning alone, this is a technology that could save hundreds of billions of pounds, while speeding up processes and keeping humans out of harm’s way.

But Professor Kell added:

We haven't really had anything since on that, from a UK government funding perspective.

This is a national shame, especially when such excellent progress had been made in the sector, before the pandemic struck.

Where the Industrial Strategy identified robotics as one of the ‘eight great technologies’ critical to Britain’s future economic prosperity, Kell explained, it is now “just a sub-bullet” under AI in government documents. (The technologies are linked, in the sense that robots can provide data to AI systems, or be controlled or informed by them, but they are best regarded as separate disciplines.)

Despite these frustrations and lost momentum, there is still much to be optimistic about in UK robotics, the Professor suggested. While the UK is, according to the International Federation of Robotics, a long way down the list of nations that are automating industrial processes – Britain has a low robot density, compared to other industrialized nations – in itself this is misleading, he said:

It's a bit of a blunt metric, because the UK is really, really good at customizing things for one-off applications in the kind of dirty, dangerous places where you really can't go. That's something that the UK excels in. 

So, I don't get too worried about those [statistics], because they are really about the mega-factories with huge production lines and masses of robots. I have quite an optimistic outlook on what the UK is able to offer the rest of the world in terms of robotics.

As suggested above, one key area for the UK is nuclear decommissioning and dealing with the nuclear legacy – both of spent nuclear material from power-stations and, separately, from weapons that have been taken out of service. The former is a critical focus for Sellafield – which has been in the headlines recently for security failings at the site.

That aside, Dr Melissa Willis is Robotics Research & Manufacturing Research Lead for Sellafield. She explained why these technologies – especially autonomous systems – will be so important economically, environmentally, and in terms of human safety:

Certainly, the biggest thing for us would be that it could potentially accelerate our decommissioning programme. At the moment, all of our robotic systems are tele-operated. So, there is still a human in control at the other end. They're sat in a nice, safe, warm control room, but ultimately, we're limited by the hours that a human can work. 

And we have areas where we are doing highly repetitive tasks and inspections [which risks error]. If we can start automating those systems properly, full autonomy would free up human operators to do something else, letting the robot continue its task. 

So, for us, we could almost double our workforce [in that way], which would massively accelerate how fast we can decommission Sellafield.

These are not minor issues: the site’s full decommissioning timeline currently stands at over 100 years, at an annual cost of £2-3 billion. That means a total projected cost of up to £300 billion, but only at current rates. In 100 years, who knows what that figure might be – £1 trillion, perhaps, for decommissioning a single facility?

But even that only represents a fraction of the economic challenge, Dr Willis implied:

There are 16 other sites across the UK that also need to be decommissioned. Sellafield is often seen as the biggest and nastiest, so if it works for us, then generally it should work for the other nuclear sites as well. 

So, if we can decrease that timeline [using robotics], then it's a better use of taxpayers’ money. Not only would it help Sellafield, but also the whole nuclear decommissioning industry.

Looked at in this light, robotics offers huge potential for the UK economy, both in terms of growth and jobs, and colossal cost savings. But she added:

For us, the biggest driver behind using robotic systems is taking people away from harm. We've got areas onsite that are dangerous, dirty, and dull. They're environments where you wouldn't want to put a human. Plus, we've got areas that are so high in radiation, you can't physically put a human in there. So, the only option available to us is remote decommissioning, and robotics is paves the way to that. 

Operator safety is our number-one priority and our biggest driver. So, yes, the cost benefits are nice, and the reduction in decommissioning timelines is a great benefit, and that's probably what will sell it to the government. But actually, our day-to-day operation is all about what we can do to make our lives easier and safer on the site.

A path to full autonomy

However, the techUK event was specifically about autonomous robots: machines that can carry out missions independently, without human control. 

This is an area about which doom merchants are especially fearful. But there are many good reasons for developing safe autonomy. Among them are the avoidance of fatal human error – the core reason for the advent of autonomous vehicles, for example. And another is the communications time-lag that inevitably occurs over long distances, due to that universal limit in physics, the speed of light (including radio waves).

At interplanetary scale, the lag might be several minutes – from Earth to Mars, for example); but even on Earth it might be significant enough to make real-time human control of a robot dangerous in some environments. This is especially true under the sea, because radio waves propagate poorly in salt water. 

Therefore, robots that can carry out tasks autonomously – without tethers or tele-operation – may actually be safer in many instances. So, how far are we from reaching that advanced stage of development for extreme environments? Especially ones as dangerous as nuclear decommissioning, where trust in an autonomous machine would need to be absolute?

She said:

That's not an easy question to answer; it depends on a lot of factors. But I will say that in our robotics roadmap, we see full autonomy from 2040 onwards. Nuclear is always a bit slow at taking on and trusting systems. But at the same time, we are one of the leading places where we can use those systems. So, my guesstimate is in 20-plus years’ time, but I won't bet my job on that!

But that would only be achievable with sustained investment and strategic support.

Another challenge in the adoption of robotics in extreme or hazardous environments is the devices’ security – concerns heightened, in general terms, by reported cyberattacks and vulnerabilities at Sellafield. It stands to reason that an autonomous, semi-autonomous, or even tele-operated robot could be extremely dangerous in the hands of a hostile state, bad actor, or opportunistic hacker.

Nicholas Zylberglajt is co-founder and CEO of Unmanned Life, a specialist developer of autonomous fleets or swarms of robotic systems, including drones. He said:

Our whole approach is fundamentally based on cybersecurity. Two weeks ago, we were in Malaysia, exhibiting with PETRONAS and Cisco, how we integrated Cisco cybersecurity products into ours.

Another example – without going into the technical details – is our network approach. We are able to integrate many different types of networks. So, now we are deploying with one of the energy utilities leaders in Europe, also present in the UK, and deploying at nuclear plants in Europe. 

All these private networks, inverted proxies, etcetera, this is how you protect from the cyber threat. Very highly protected networks that are private, that are very difficult to penetrate. But also, very difficult for robots, because you then have the difficulty of discovering the robot [on the network]. That's where you need to have software that is very good.

Professor Kell added that, for Jacobs:

Security is always a challenge, but it's something that we address on a case-by-case basis. I work with a wide range of different clients – space, defence, utilities, road and rail networks. And the British Army can have very different requirements, challenges, and risks to Thames Water!

My take

Indeed. And let’s hope that the government sees the opportunity to set out a new, comprehensive Industrial Strategy. One that recognizes anew the potentially huge contribution that robotics can make to the economy – and to human safety.