en.Wedoany.com Reported - WSP Artificial Intelligence (AI) explains how the organization is leveraging digital machine learning tools to reshape the maintenance of UK infrastructure. The core principle is that AI should not replace engineers, but rather handle large-scale tedious tasks that humans cannot manage.
Dan Scott, Chief Data Scientist and AI Lead for WSP UK and Ireland, along with James Montrose, Head of Digital Consulting, described how the consultancy uses AI to digitize decades of paper documents and accelerate daily task completion. WSP's AI strategy encompasses a dual approach: building AI capabilities within its own workforce while collaborating with clients to optimize workflows. This includes partnerships with Microsoft and prototyping tools through a dedicated AI accelerator to drive successful pilot projects into practical application.
Scott stated that WSP rolled out the Microsoft Copilot product to UK employees approximately 18 months to two years ago, continuously evaluating its adoption rate and benefits. By using time codes in consulting billing, the company tracks time spent on non-project activities such as administrative work and meeting minutes, finding that employees who use Copilot more than once a month save an average of about 165 minutes per week. WSP is currently piloting a more tightly controlled system that provides similar functionality while managing potential data risks associated with AI. Internal AI uses include automating meeting minutes, administrative tasks, and simple document searches. Scott noted that these low-cognition, time-consuming tasks represent the "sweet spot" for current technology, as the time required to complete them accumulates and frustrates engineers.
In the rail sector, WSP is collaborating with Network Rail to build an AI foundation for predictive maintenance and long-term investment planning of aging assets. Both parties currently use AI tools to sift through vast amounts of asset-related data, including scanned reports, images, and handwritten notes stored in SharePoint or off-site warehouses. Discussing this work, Scott asked: "Is it possible to use AI to predict which parts I need to repair and when?" He explained that the reality is that AI must first be used to fix the underlying data, bringing it into a structured state, before such conversations can begin. The technology developed by WSP converts scanned documents into accessible data and links historical records, real-time operational data, and asset management systems. In defect management, Scott described the complexity of the current process: someone identifies a defect, quantifies it, proposes a repair, then a work order is created and delivered, and the work is completed. However, this chain is difficult to trace because it has long been managed solely through written reports from field engineers. Additionally, WSP combines asset failure records with UK Met Office climate data to build predictive models that determine when and where assets will fail under extreme weather conditions. Montrose explained that the model supports operational expenditure and capital expenditure planning, citing a recent disruption in London caused by high temperatures as an example.
In the water sector, WSP developed an AI assistant project called Wisdom for Northumbrian Water. This AI agent answers operational questions by integrating disparate information from company resources, such as identifying a pump's site location, displaying its current output and historical performance, and retrieving maintenance records and design specifications. Scott noted that Northumbrian Water faces significant capital investment needs over the next five to ten years, coupled with an aging workforce where much operational knowledge resides with employees nearing retirement. The Wisdom system can be seen as a ChatGPT for conversing with assets. Beyond cleaning and querying historical records, Scott also described running millions of combinations of expenditure and stress factors through AI to explore complex problem spaces that humans cannot understand alone, identifying strategies and solutions for the water industry.
Montrose emphasized that the core of using AI is to elevate everyone's value chain, fostering better conversations and understanding impacts, rather than simply being more persuasive than others. Scott reiterated the engineering firm's principle: never use AI to do something you cannot do yourself. This principle guides how WSP allows engineers to use AI tools, including promoting programming training for employees to ensure engineers can understand and verify AI-coded applications.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
