2023 Best Use of Technology: Artificial Intelligence

Optimising the HS2 programme

Alice Technologies

SCS Railways (Skanska Costain Strabag Joint Venture) is delivering the London Tunnels contract for HS2. Extending from Euston Station to West Ruislip, 21km of twin bore Tunnel Boring Machine (TBM) tunnels, eight shafts and 5km of Sprayed Concrete Lining (SCL) tunnelling will be constructed.

AssetScan

CC Informatics

CCI have developed a deterioration and defect identification and tracking AI algorithm. AssetScan is a trademarked and patent pending AI which has been used to assess the requirements for maintenance at a large masonry faced concrete gravity structure in Wales. The AI uses photographic data to detect deterioration in construction media such as masonry, concrete, cladding and brickwork. The AI was used to show where repair works had been undertaken and where further maintenance works would be required on the downstream face of the structure. AssetScan was used in combination with a detailed drone survey of the dam. The photographic data was then used to create bespoke training data which was used to teach the AssetScan AI to detect issues with pointing. DCWW have used this information to inform the effort and materials required to finish repointing the dam. The project has won an internal DCWW health and safety award.

Safe Dig AI

Morrison Water Services and Zensar Technologies

Jointly developed by Zensar and Morrison Water Services, Safe Dig AI is an AI-led Robotic Process Automation (RPA) platform that rapidly processes utility map data. The solution provides us with accurate safety insights via a user-friendly app. This is an entirely innovative use of AI technology which improves efficiency and safety on site and unique to the Water Division. We believe it has much to offer the engineering and construction sectors in terms of safety and efficiency. It also eliminates the need for operatives to juggle multiple paper maps, reducing not just carbon footprint, but the likelihood of human error.

AIVR Machine Learning

One Big Circle

Spurred by the challenges of attending site throughout the Covid-19 pandemic, One Big Circle’s AIVR (Automated Intelligent Video Review) system was developed to capture and transmit critical video data from the rail via train borne cameras, to make data rapidly and easily accessible via an online platform. AIVR enables the rail workforce to remotely inspect project sites, accessing a variety of Forwards Facing, Thermal, OLE, and Line-scanning video; AIVR empowers industry workers to reduce boots on ballast and complete survey and monitoring activities remotely and safely. To enhance the remote inspection benefits of AIVR, One Big Circle has developed a range of intuitive Machine Learning and AI models within the platform, to assist users in drawing further insight from AIVR data. AIVR’s AI capabilities automatically detect rail assets, thermal faults, and patterns of behaviour within the platform, to assist users to draw further insight for intelligent asset management.

Smart environmental site monitoring and digital community engagement

UBY

UBY is a new monitoring technology provider with a mission to make construction leaner, safer and more efficient. Furthermore, it is to improve and protect the wellbeing of residents that surround urban developments. We achieve this by being different. That is, we use AI enhanced sensors and patented algorithms and software to provide useful and actionable information in an easily understandable format. Feedback from Leonard Davoux, Bouygues UK (Hallsville Quarter Phase 3). "We were the first construction project in the UK to install UBY's smart noise sensors and within weeks we'd reduced the time spent on responding to resident complaints and sensor alerts. The technology totally changed and improved how we were able to respond and meant that we weren't wasting time trying to investigate and resolve each occurrence. This allows us to then get on with the main tasks in hand, which is critical on any major project".

Daisy

WSP

Daisy represents the performance-based design processes of the future. Powered by AI and machine learning, it can analyse a building's performance across multiple metrics, such as carbon emissions, energy consumption, human comfort, economic returns on investment, and environmental criteria. Daisy creates thousands of models using a genetic algorithm, testing them against multiple objectives, to show us the most efficient option. By considering the complex interactions and trade-offs between these objectives, we can optimise positives and mitigate damage. This shift from "business as usual" sustainable design to transformative regenerative design is what Daisy aims to achieve. The tool's development involved a rigorous research and development strategy, including a proof-of-concept prototype, dataset creation, machine learning techniques, and optimization of a benchmark building. Daisy's methodology has shown significant improvements in energy use intensity, and its scalability and innovation set it apart as a groundbreaking solution for reducing carbon emissions and enhancing building.