Is AI the answer to our maintenance woes?

“When I grow up, I want to go into maintenance,” said no child ever.

Maintenance is critical for keeping our vital infrastructure in peak condition. But, when set against the delivery of new assets, it can appear like the poor, less exciting, relative.

Yet, a lack of it can have serious consequences: putting an asset out of use, prompting it to malfunction, or worse still, causing a public tragedy.

Today though, those in charge of maintaining our critical infrastructure have a new tool in their armoury: artificial intelligence (AI).

AI is changing whole industries

The benefits and potential of AI across multiple areas of our lives have been well discussed.

In sport, AI, we have been told, will help to pick the best players and develop the most effective team tactics.

In the office, it is summarising vast, unenticing documents into crisp summaries and in healthcare, helping medical professionals to forge their ways to breakthroughs.

But in infrastructure maintenance too AI - coupled with rigorous information practices and high-quality data - is also making its mark and changing how we work for the better.

Potholes and predicting maintenance

The technology’s potential in helping to manage and maintain our infrastructure was recently underlined by the UK prime minister. He hit on AI as the tool to tackle his manifesto pledge to fix an additional million potholes a year.

But this is not some future utopia.

AI can - and already is - providing infrastructure operators with the ability to better monitor and, through simulation modelling, intelligently predict, asset health and maintenance needs.

How AI is making maintenance safer

One area in which AI is having a big impact is safety. How? By literally keeping maintenance teams out of harm’s way.

Sund& Bælt, which maintains some of the world’s largest bridges, has explained how this works in practice.

Traditionally, its maintenance regime would include mountaineers, hooked up with safety equipment, scaling their bridges to inspect for cracks or defects.

Today, it uses drones – a process which can take one day instead of one month.

The data captured, alongside other relevant data, is ‘read’ by powerful AI tools to provide valuable knowledge to the teams.

And instead of scheduled maintenance every few years, it has, thanks to AI, a ‘live’ second by second picture of the bridge condition.

AI in highways

AI is also playing an increasingly significant role in the safety of the highways network thanks to the Bentley Systems company, Blyncsy.

Blyncsy recently published a comprehensive public map of all interstate highways in the continental United States showing their key highways assets.

The map was developed using crowdsourced dash camera imagery from over one million vehicles.

Coupled with Blyncsy’s powerful AI image analysis toolset, it can detect over 40 different road conditions and issues in near-real time.

The data is available to all state U.S. Department of Transportation agencies to help them better address safety, maintenance, and repair operations.

Data for digital twins

The value of tools like building information modelling (BIM) and digital twins is multiplied when they’re powered by AI.

AI algorithms can continuously learn from the data collected by digital twins. This helps to improve their accuracy and ability to predict over time, enhancing their value for infrastructure owner operators.

Why a sound data strategy is essential

The infrastructure intelligence that AI can offer depends entirely on the existence - or establishment - of a quality data foundation and a sound master data management (MDM) strategy.

This will ensure the uniformity, accuracy, stewardship and accountability of any organisation’s data.

Without this solid foundation for implementing AI, its potential cannot be realised.

How to manage maintenance data

Asset maintenance teams looking to get their data in order ahead of embracing AI, should ask themselves these questions:

• What insights do I need?
• What datasets do I need to integrate to get these insights?
• Can I be confident that my data can be trusted?
• Will it enable better value in performance?

Waving goodbye to maintenance woes

Infrastructure asset failures – and everything that goes with them in terms of downtime, safety, time and cost – can become a thing of the past.

Digital and AI can drive a new generation of leaders in asset management who are proud of the contribution they can make to society.

And future children may meaningfully say: “when I grow up, I want to go into maintenance”.