Is robotics the key to repairing societal infrastructure?

Anyone who’s been stuck in a gridlock during road repairs knows that maintaining our essential infrastructure can disrupt daily life for millions of people.

Repairing tarmac, rail networks, gas or water mains, cables, sewers, street lights, traffic control systems... the list goes on at ground level or beneath it. Meanwhile, fixing bridges, towers, and wind turbines presents its own set of challenges.

In crowded cities, disruptions can be long-lasting, expensive and frustrating. More, they often reveal how inefficient our systems are. The economic impact can be massive, and yet our infrastructures demand continuous building work, inspection, repair, renewal, and upgrading in order to prevent a worse type of disruption: critical failure.

Fixing these problems swiftly and efficiently could be a transformative application of robotics, according to a new white paper from UK-RAS, the umbrella organisation for British robotics research. Robotic and Autonomous Systems for Resilient Infrastructure is a UK analysis, but with conclusions and recommendations that have global applicability. As UK-RAS notes:

Our vision is of a society where infrastructure engineering is undertaken with zero disruption to human activity and zero environmental impact.

The first question is why these problems occur in the first place. The white paper explains that the real disruptions are rarely caused by the maintenance work itself, but by creating the safe access mechanisms for human beings to do the work. Workers have to dig trenches, erect scaffolding, build gantries, or close roads and rail lines before essential repairs can take place. Building these systems often takes weeks for a few days – eventual – work.

The paper explains that robots could help by stripping away these parts of the process:

Robotics and Autonomous Systems (RAS) can operate in dangerous and challenging locations, such as inside underground pipes, or at height underneath bridges, or on live roads to perform inspection, repair, maintenance, and removal tasks.

Standard processes in maintenance engineering include: pre-investigation; site surveys; monitoring; construction and creation; maintenance; repair; and dismantling and disposal. In many cases, these can be automated or carried out by robots: autonomous bridge inspections, water pipe repair bots coordinated by autonomous planning systems, and so on. Meanwhile, the need to build scaffolding to inspect a building (let alone repair it), could be done away with by bringing in a drone.

The white paper says that robots can positively disrupt the engineering sector by:

Improving speed, quality and timeliness of infrastructure engineering, while reducing direct costs, economic impacts, material waste, energy usage, environmental damage, and risk to human operators.

The cost element alone is massive. In the UK, for example, the government has committed to spending £500 billion by 2020-21 on traditional infrastructure projects.

Central investment in national networks is critical for accruing a broad range of societal benefits, such as: supporting growth and creating jobs; raising productivity; driving efficiency; and boosting international competitiveness. In each of these areas, RAS can play a significant role, says the white paper.

Getting smart

It's an attractive picture, but it isn’t complete. Our reports from CERN earlier this year found the missing piece of the puzzle: smart asset management, and the embedding of intelligence into the infrastructure itself.

As those reports revealed, the CERN campus is a smart city with millions of physical assets. Some are vast or exist in environments that are hazardous to humans, while others are minute or inaccessible – a single bolt hidden somewhere in the 27km loop of the Large Hadron Collider. Without smart asset management, big science experiments would need to be shut down in order to locate and fix tiny problems (and much larger ones).

Similar challenges apply in any town or city. So the solution is for the component itself to tell you where it is, what it does, what it connects to, and when it needs fixing – and for its maintenance or obsolescence to be planned for in advance. Once on the system, much of this can be automated. The white paper concurs:

The convergence of RAS with the ability to embed intelligence into assets, extract intelligence from big data, and the development of new manufacturing materials and processes has created a platform to overcome the complex engineering challenges associated with installing and maintaining infrastructure networks.

The opportunities here are vast, it suggests:

The companies that successfully deploy and exploit RAS will be at the forefront of the fourth industrial revolution. In addition to providing an unparalleled customer experience, it is likely that these firms will unlock value chains that generate completely new economic activities.

Bold claims. But is anyone doing it for real?  Patches of innovation are certainly emerging. In the UK, for example, there are a long-term plans for robots to replace diggers in Leeds and turn it into the world’s first self-repairing city. Meanwhile, Oxfordshire County Council is using AI and predictive analytics to provide early warnings of travel problems in the region, so that the right resources can be mobilised.

But in the UK, one obstacle to widespread robot adoption – including driverless vehicles and drones – is legislative. The UK is small and densely populated, its skies and roads are crowded, and its aviation sector is (understandably) cautious. In the bigger skies of the US, soft-touch regulation is clearing the way for maintenance drones.

Then there are the employment impacts. The UK’s engineering workers would resist mass automation. Currently, robotics is more closely associated with job losses than job creation, and there is ample evidence of its negative impacts. For example: out of more than 2.5 million jobs in UK real estate and construction, 34 per cent are at high risk from automation, according to Deloitte, and 16 per cent are at medium risk. Together, that’s half the workforce.

However, UK-RAS contrasts this with the US automotive industry. Between 2010 and 2015, car makers installed more than 60,000 industrial robots, and the number of human workers rose by 230,000. If true, this suggests that a single robot can create up to four new jobs.

But in infrastructure maintenance and engineering, would these be menial support roles, or something more? The white paper says:

Infrastructure creation and operation has traditionally provided work for many of the lower skilled in society. Trends towards robotics will demand more high-level skills from the ever-reducing number of people employed in the industry. A trend towards increased use of robotics will make those vital industries more attractive to the tech-savvy youth of today, providing the technological stimulus to attract the next generation of engineers and new startup companies.

So the report is saying is that robotics will strip away low-skilled jobs and replace them with skilled, high-tech opportunities – in the engineering sector, at least. That deepens social inequality, with automation and low-wage roles locked in a fatal embrace, as UK-RAS acknowledges.  The white paper urges that robotic innovation should “free up our workforce” to:

Tackle the more complex, creative, and challenging issues facing our ageing infrastructure. The industry is desperately short of workers, particularly at higher skill levels. Reducing the unskilled workload will give the industry the opportunity to retrain existing workers rather than recruit from abroad.

And post Brexit, of course, the UK at least may have no choice...

But elsewhere, these societal divisions are less clear. In some sectors of the economy – in banking, finance, the media, legal services, publishing, and healthcare, for example – automation, AI, and robotics are replacing professional roles, at least at entry level. So robots are knocking out the first rungs of the middle-class career ladder, too, making it harder for newly qualified people to succeed.

My take

Societal challenges aside, where does the robot revolution offer when it comes to its critical national infrastructure?

Most Brits would tell you that the UK is uniquely susceptible to infrastructure problems and inefficiency: railway engineering that seemingly lasts for decades; coned-off motorway lanes with no workers in sight; months of scaffolding masking a week’s exterior decorating, and so on.

UK-RAS, meanwhile, says that the UK is well placed to capitalise on the opportunity for exactly the same reason, combined with the density of its national infrastructure, and its deep pool of engineering talent.

Either way, these changes are coming. So let’s hope that the government increases its central investment in RAS technologies [see our UK Robotics Week reports, passim] so the country can be in the locomotive of new industrial opportunities, creating new jobs and new economic growth.

The alternative? Sitting at the back of the train, grumbling about delays.