After Prime Minister Keir Starmer announced the Government’s AI ambitions this week, Russell Goodenough, head of AI for CGI in the UK and Australia, discusses the benefits AI could bring to the UK’s public sector.

Pioneering AI for infrastructure maintenance

As well as being a huge week for Government and the tech sector, today (15th January) also marks an important day: National Pothole Day. The day specifically aims to raise awareness of the impact of potholes on cyclists, which is a clear frustration and indeed danger for all road users.

For many, the idea that AI could help reduce the maintenance burden on local government teams striving to locate and fix potholes might seem ambitious – particularly given the complexity of identifying and prioritising issues across extensive networks of roads and infrastructure. However, this is a use case that is already within reach.

One example of AI’s potential in this area is Machine Vision, an industrial tool being used to improve the efficiency of preventative and corrective maintenance in many sectors. Machine Vision is an advanced AI-based system that leverages machine learning and edge computing to interpret visual data in real-time. Rather than have employees comb through hours of videos and endless images, the tech itself can analyse the source materials to identify infrastructure issues like potholes or equipment failures.

How AI Technology Works

By using advanced transformer model architecture, which is the same foundational technology behind 'Large Language Models' like ChatGPT, Machine Vision can analyse static and video imagery to assist in maintaining transport infrastructure. This activity combines edge AI and machine learning to interpret videos and images in a human-like way, enhancing asset and infrastructure monitoring across various industries.

The system operates by capturing CCTV data (and other IoT sensors), processing this data at the point of capture using edge computing and generative AI, and then integrating the processed information into data operations. This approach reduces latency, decreases cloud storage and data transfer costs, and enables real-time analysis tailored to specific objectives. It means that local governments can prioritise maintenance tasks effectively and reduce overall operational costs whilst improving citizen experience.

Adaptability and practical applications

A key advantage of this approach is its adaptability. For instance, integrating with existing CCTV systems means there is no need for significant additional investment in hardware. AI systems can process video feeds and public-sourced images to identify and assess the severity of potholes, providing actionable insights to direct maintenance teams effectively, ensuring they have the right tools for the job.

And this technology is not just theoretical, Machine Vision technology has already been applied in various scenarios to improve efficiency and safety. In the rail sector in the UK, it’s been applied to enhance existing alert systems and provide real-time monitoring of remote and unattended sites. Additionally, it can monitor uncrewed facilities such as water treatment plants, dams, pipes, and substations, providing alerts for stopped or slow equipment, leaks or unusual activity. It also means rather than sending rubbish collection crews on a fixed basis, collection schedules could be dynamic based on the exact levels of rubbish in bin stores – all thanks to the power of AI and sensors.

By replacing periodic manual inspection with continuous automatic analysis, local governments that engage with this technology can adopt predictive and proactive operational models. This leads to increased data quality, improved real-time data analysis, and the ability to capture more data points without increasing costs. Ultimately, it supports these organisations to improve the reliability of their key assets as well as supporting with sustainability, safety and regulatory compliance.

Broader applications and benefits

Local governments operate with limited resources, making it essential to focus efforts where they’re needed most. AI tools like these can bring significant advantages, from reducing costs through better resource allocation to boosting public satisfaction by addressing infrastructure issues more efficiently. They also enhance safety for communities by ensuring that critical maintenance tasks are prioritised and executed effectively.

Whilst some may say deploying AI to fix a pothole is like using a sledgehammer to crack a nut, the decision to utilise AI and other tech is a huge moment for many organisations. It might mean potholes today, or bins tomorrow, but exploring how AI can complement existing systems is an important step in modernising public services and addressing long-standing persistent challenges efficiently. Finally then, we may crack the challenge of meeting citizen needs with a clear eye on the public purse at the same time.