On the factory floor, machine downtime is a disaster – production can slow to a crawl or a complete stop, wasting precious time and money. To dodge downtime, manufacturers typically plan rigorous maintenance regimes – fixing problems before they even occur. That’s great for keeping production running at full pace, but costly in terms of having the right experts and spare parts to hand at all times. Preventative maintenance is labour intensive and expensive. The alternative? Have faith in data – and artificial intelligence.
The acceleration of smart manufacturing means that an incredible amount of data now flows from the factory floor. Hidden within that deluge is, potentially, everything a manufacturer needs to know about every machine under their control. And to make sense of that data, many are turning to artificial intelligence. But, in the pursuit of zero downtime, this creates a new kind of risk: are manufacturers willing to place their trust in AI to make the big calls about maintenance and safety?
“It’s one thing to have an algorithm tell you which machine to maintain next, but it’s another thing to change your whole maintenance strategy and to then rely on it,” says Alex Holland, senior director and new venture lead for Social Innovation business at Hitachi Europe. “What if something changes in your operational environment that results in the algorithm no longer being so reliable? It comes down to this idea of both trustable AI, and also new approaches to risk management.” Sure, an AI, once fed on a feast of data from the factory floor, might be able to design a better maintenance schedule than any human – but can it be trusted? Is it worth the risk?
To help the manufacturing and transport industries feel more confident in embracing artificial intelligence, Hitachi has taken an unusual step: it’s offering AI insurance. Working in partnership with Swiss Re Corporate Solutions, the commercial insurance arm of reinsurance group Swiss Re, Hitachi is essentially putting its money where its mouth is: put your trust in our AI to run your factory maintenance plan and you, the manufacturer, are insured against the AI making a mistake. The ultimate aim is to evaluate risk more precisely, using machine-sensor data and continuous diagnostics across the industrial sector to ensure pay-outs are made automatically and effectively.
“The huge benefit that I see here is that we start to explore areas that have been uninsurable because of the lack of data. Now, with data becoming available, we can really start to analyse it and we can look to offer new products,” says Stefan Wunderlich, internet of things proposition lead at Swiss Re. For insurers, this wealth of data – and an AI smart enough to make smart decisions based on it – allows it to take on risk that it was previously reluctant to have on its balance sheet.
WIRED in conversation with Alex Holland and Stefan Wunderlich
The partnership between Hitachi and Swiss Re remains in its early stages, but Wunderlich has already had encouraging conversations with manufacturers looking to make better sense of data that they have been collecting for some time. “Customers are more and more willing to open up their books and say, ‘We have this kind of data, if you need it to take off some risks, we understand,’” Wunderlich says.
Hitachi has been in the business of predictive maintenance for several years through its Lumada internet of things data insights platform. But encouraging manufacturers – many of which have very set ways of working – to fundamentally change how they operate and listen more to the data the factory floor is already producing has proven challenging. “Organisations are interested in the technology, they’ll even buy the technology, but to change their whole approach to maintenance is sometimes significantly more difficult,” says Holland. And that’s where Swiss Re comes in. The insurance firm provides what is known as a parametric warranty for Hitachi’s AI maintenance system. This is a warranty for Hitachi’s prognostic system that will pay out if it makes an incorrect prediction.
The prognostic system might show, for example, that a certain machine is not going to fail in the next few weeks or months, so a manufacturer decides not to maintain it. “But then say it does fail – there’s a payout from the warranty,” explains Holland. “This gives the end customer not only some kind of compensation when it fails, but actually, more importantly, because these failures are very rare anyway, a greater reassurance and a way of handling risk for them.” By insuring against the AI making a mistake – even if it is rare – Hitachi is hopeful that more companies in the manufacturing and transport sectors will change their maintenance strategies to make use of smart, insightful data analysis. Such a strategy, Holland explains, could reduce maintenance costs but also improve machine uptime.
This shift to data-led, AI-based decision-making could further accelerate another trend: equipment as a service. With the internet of things and artificial intelligence providing ever more accurate data and insights, the onus on keeping machines on the factory floor shifts towards original equipment manufacturers, or OEMs. “The move towards equipment as a service has potentially significant benefits for OEMs because they get to capture a larger part of the after sales market,” says Holland. For manufacturers, this provides peace of mind and a better level of service. For OEMs, it provides a steady stream of revenue throughout a machine’s lifetime. “IoT is really opening up an entire new world of insurable products and spaces,” says Wunderlich. “It’s expanding the insurability of risks to a totally new level.”
Holland is hopeful that the protection of insurance can persuade more companies to take the plunge and start leveraging the benefits of the internet of things and artificial intelligence. “This technology is proven and available today,” he says. “We’ve started by looking at factories, but this could be used in any industry that relies on complex, critical assets where downtime is hugely disruptive and expensive.” From transport networks to utility companies, insuring AI could unlock huge benefits for big industry and consumers alike.
Modern life is saturated with data, and new technologies are emerging nearly every day – but how can we use these innovations to make a real difference to the world? Hitachi believes that social innovation should underpin everything it does, so it can find ways to tackle the biggest issues we face today. Visit Social-Innovation.Hitachi to learn how Hitachi Social Innovation is Powering Good and helping drive change across the globe.