How Network Rail is saving millions by using machine learning - New Civil Engineer

Cost savings of up to £30M were identified on Network Rail's Great Western project through use of machine learning and wider construction industry benefits are set to be unlocked by a new investment from a Google venture business. Network Rail leveraged the savings by using construction risk forecasting platform nPlan, which has just been boosted by a £13.5M investment from GV (formerly Google Ventures). According to nPlan, wider use of machine learning that this investment will support could save the global construction sector almost £730bn each year through spotting delays and recommending improvements with an accuracy and scale previously not possible. The firm has said that such data-led insights would effectively reduce the volatility of and increase investor confidence in construction projects. By using some of the most powerful machine learning capabilities in the world to analyse what worked and what didn't in past projects, we can help our customers work out what's going to derail their own initiatives, and stop problems happening before they even appear." Speaking about the due diligence undertaken by GV ahead of the investment, GV general partner Tom Hulme said that he was "blown away" by the benefits organisations such as Network Rail saw in applying modern machine learning techniques to such a difficult analogue problem".