BTS | HARDING PRIZE COMPETITION 2023
PROGRESS USING AI
FORECASTING TUNNELLING
The Harding Prize Competition is named after the founder of the British Tunnelling Society, Sir Harold Harding, and is for entrants under the age of 33. Shortlisted papers were presented to the BTS evening meeting on 13 April, at the Institution of Civil Engineers, in London.
Jake Rankin, data scientist with J Murphy & Sons Ltd for this project,
was voted winner of the 2023 Competition with a paper entitled ‘Artificial Intelligence for Forecasting Tunnelling Progress’. It is published here with the author’s permission
ABSTRACT Tunnel boring is essential for many construction projects, yet it is challenging to accurately forecast productivity due to many influencing factors including ground conditions. Whilst ground survey data has improved over the years, significant gaps will always remain due to the inability to carry out boreholes for every metre of the tunnel. Accurate forecasts are essential for project management and can help to identify potential excavation challenges, manage stakeholders, and ensuring a confident forecast. Murphy Group engaged me to develop and apply
artificial intelligence to tunnel boring, specifically to forecast tunnelling progress. The algorithm selected for this study was a neural network, which is an algorithm that can learn patterns from real-world data that are non-linear, complex, and incomplete. Such
data are characteristic and prevalent within soil-tool interactions, making neural networks a case study worth investigating. The work in this paper shows two case studies. The first was on a prior project – Southwark Tunnel,
in London – which provided an array of machine sensor data. The project data would help in this first case study by enabling understanding of how well neural networks can capture tunnelling performance and machine behaviour, based on the performance data from the tunnel boring machine (TBM) on that project. The second case study would take forward
development of the neural networks by further applying the system to a live tunnel project, the North Bristol Relief Sewer (NBRS), and this time by only using geological data. This approach was designed to mimic
Above, figure 1: The EPB machine used for the project ALL IMAGES COURTESY OF J MURPHY & SONS LTD, UNLESS OTHERWISE REFERENCED
32 | August 2023
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49
