Mehdi Mahmoudysepehr: An engineer in Geomechanics and PhD student at Chaire Construction 4.0

5 minutes of reading

‘To better understand and interpret the behaviour of tunnel boring machines in order to anticipate and solve technical problems within the industry is the subject of Mehdi Mahmoudysepehr’s doctoral research project. He is a PhD student at Chaire Construction 4.0. He tells us more about it below!


Your thesis is entitled ‘Modelling the behaviour of tunnel boring machines and their impact on their environment.’ Could you tell us what it is about please?

The aim of my thesis is to improve knowledge around the behaviour of tunnel boring machines during excavation work, and, in particular, to find potential correlations between the operational parameters of tunnel boring machines, the characteristics of the terrain and the guidance.
Until now, analytical methods have been used to guide tunnel boring machines. We now want to improve our calculations by analysing the large volumes of data collected in the field.

Better knowledge of the relationship between internal and external forces experienced by tunnel boring machines can then help engineers and experts improve current calculations and anticipate and solve technical problems.

At the same time, I am interested in data-driven approaches which use Big Data analytics to help make strategic decisions. Using these methods, it is possible to model the behaviour of tunnel boring machines and improve operations such as guidance.

Would you like a concrete example? The skirt (rear part of the shield) can sometimes get stuck when a tunnel boring machine is being driven. This situation can be anticipated and avoided by improving knowledge about the balance of forces being exerted. Another example is that with a better knowledge of the terrain, performance can be improved by accelerating the speed of advance.


Mehdi has a Master’s degree in Geomechanics, Civil Engineering and Risks from the Université Grenoble Alpes. He also has a degree in Mining Engineering (mining extraction) from the University of KASHAN (Department of Engineering) in Iran.

Mehdi began his doctoral thesis at LaMcube (Laboratoire de Mécanique Multiphysique et Multiéchelle) at the ‘Ecole Centrale de Lille’ withinChaire Construction 4.0 under the guidance of Professor Zoubeir LAFHAJ and Professor Amine AMMAR from the ‘Ecole Nationale Supérieure d’Arts et Métiers’ in Angers. Within Bouygues Travaux Public, Mehdi’s thesis is a part of ‘Tunnel Lab’, which is managed by Nicolas BRAUD. Julien Larseneur is a member of this team and he supports Mehdi with his work.

How have you carried out your doctoral thesis at Bouygues Construction?

My PhD started with studying available literature to identify and evaluate the analytical and numerical methods that help us acquire theoretical and technical knowledge about tunnel boring. We stand on the shoulders of giants – we always rely on the work done by those before us to enable us to go further!

I conducted a study to improve our knowledge about the balance of forces exerted on tunnel boring machines in order to better understand and interpret their behaviour. To achieve this, new sensors were installed on a tunnel boring machine in the Grand Paris Express.

After that, I was able to get down to business! In order to model the behaviour of the tunnel boring machine, I used data-driven methods to evaluate the operational and trajectory parameters of the tunnel boring machine, as well as the geomechanical characteristics of the terrain. The aim was to find any significant correlations between these parameters.
This assessment of the tunnel boring machine’s operational parameters gave us the skills we needed to better understand the physics of what is happening during excavation.

Finally, I developed an approach based on the instruments on the tunnel boring machine. This involved deploying new sensors together with my electronic technician colleagues on the team to calculate the loads exerted on the skirt structure. This method improved the calculation of the balance of forces. As a result, it could be used to improve the calculation of the forces undergone by the tunnel boring machine and to solve technical problems relating to its progress.

The final stage, which is still in progress, now consists of using statistical learning methods (Machine Learning and Deep Learning) to model the behaviour of tunnel boring machines and improve their guidance.

What is next?

This study may be a first step towards using ‘learning’ algorithms to automate the guidance of tunnel boring machines, or at least to help the pilot to drive a tunnel boring machine with more information and therefore improve reliability and safety.

One of the next objectives was to try it out on new construction sites. To date, only one tunnel boring machine has been equipped with sensors, and, due to the cyber attack and stoppages on construction sites due to Covid-19, there have been some technical problems in the data feedback. My calculations for now are based on data collected between June and December 2019. However, despite these two setbacks, I am not at all discouraged – quite the opposite in fact! I am very proud that the project is still continuing, despite these difficulties.

Throughout this PhD, I have been fortunate enough to enjoy many discussions about the technical issues encountered with the experts and engineers from ‘Tunnel Lab’, ‘Direction Tunnel’ and ‘Direction Méthodes et Prix’. I will continue to do this because it is important to remain connected to what is happening day to day within the ‘business’. A PhD is not about advancing research for the sake of research alone. It is about providing, through research, concrete answers to what is required within the field.

A focus on…‘Tunnel Lab’

Tunnel Lab is in charge of transforming ideas and concepts from teams at Bouygues Travaux Publics into prototypes. These prototypes are then installed on the tunnel boring machines during production. After this crucial first step, the team’s second challenge is to work on the industrial development of these innovations so that we can offer the best services to our customers.

Tunnel Lab’s work means the team selects partners from both small and large companies to develop these innovations. The team also works in collaboration with scientists to detect and then propose new technology which meets the requirements of the business teams.

This multidisciplinary team is lead by Nicolas Braud. It is made up of experienced project managers and experts in equipment, energy, tunnels, geology, embedded computer systems and data analysis.