How is AI shaping the future of existing infrastructure?
Artificial Intelligence, machine learning and data science all have incredible potential to enhance the productivity of infrastructure throughout its lifespan. ICE, the Alan Turing Institute and TechUK all recognise the profound benefits for existing and future infrastructure, exploring the possibilities of how we adapt and build for the future.
The UK has a large body of legacy infrastructure that represents a productivity opportunity not to be wasted. Consider the use of sensors for data capture and AI to interpret patterns in the data. The ability to gather real-time data about asset condition, performance and efficiency is impacting everything from operational efficiency to predictive maintenance. This in turn enables optimised operational efficiency and longer asset lifespans.
Stakeholders can't afford to ignore this AI revolution. Instead, we need to harness the power of the AI revolution to create infrastructure technologies that deliver best outcomes wherever they're deployed.
Proximity-based data and data aggregation
Data science has the potential to transform the safety, reliability and efficiency of key infrastructure. Using AI driven pattern recognition, infrastructure failures can be predicted and pre-emptive action taken. This improves health and safety, reduces costs and reduces the risk of critical failure while delivering reliability for users.
Sectors as diverse as advertising and Rolls Royce engines already use data aggregation to gain a competitive edge. But the implications for infrastructure are huge, with a massive body of data including everything from thermal video imaging to weather satellites available to predict and prevent structural failure.
Asset monitoring is a globally available resource. The challenge will be to persuade data owners to share high-quality data freely, but this could be incentivised with the benefits of wider data sharing, resulting in lower costs, reduced service interruptions and more reliable assets.
For example, Heathrow Tunnel is a known bottleneck with a record of delays and closures. Existing systems including CCTV, sensors and call logs already capture a large amount of data but a change of approach is needed to use that data for predictive purposes. The addition of new monitoring systems could help to predict patterns that foreshadow asset failures. Planned closures could then address the problem and minimise disruption.
Capturing lost data
When people leave or retire from an organisation, they often take a considerable amount of data with them. This is rarely passed on or captured effectively when it could be a powerful tool in the rapid upskilling of support staff.
Data from wearable devices, video, audio and biometrics could all allow senior staff to narrate their work. This data would then be analysed by AI, allowing the creation of a digital twin for tacit knowledge. This would then feed back into more efficient operations and help to upskill teams. An additional benefit could be a reduction in form filling, using AI processed data for validation of work and the creation of a more skilled workforce and safer working environments.
Measures of success
So what will be the success factors when considering the use of AI in reshaping infrastructure? Apart from ROI, whole systems thinking and reliability, ICE, the Alan Turning Institute and techUK identified value to end customers and data quality as critical measures of success.
Much will depend on how infrastructure challenges are prioritised and whether the industry is prepared to adopt shared data standards to improve quality and shareability. But the possibilities for digitisation and data aggregation in infrastructure have never looked brighter.