Automation is defined as operating or controlling a process through highly automated and computerised means, minimising human intervention. Artificial Intelligence (AI) refers to the development of software and computer systems able to perform tasks that would normally require human intelligence.

Experts agree that tunnelling will be in high-demand by 2030. This is mainly due to: (a) rapid increases in urbanisation and scarcity of surface space; (b) more new commercial, industrial, and residential developments requiring services and utilities; (c) minimising the societal impact of future infrastructure projects; (d) climate change impacts e.g flooding, and (e) underground space becoming a real alternative to expanding urban space.

The need for automation: tunnelling safety and performance have improved dramatically in the past few decades, thanks to advances in mechanical, chemical and electrical fields. However, more complex systems leave less room for human error. In the case of large, state-of-the-art TBMs for example, operators must monitor numerous screens, alerts and alarms, thereby exposing them to hundreds of data points simultaneously. This demands a system to help operators interpret and act on real-time information. To achieve this, industry and academia are working to improve automation, including deploying AI that will recognise data patterns from operator-generated inputs and performance-related outputs.

The tunnelling industry is heading toward automation because it can significantly reduce costs and increase safety. Removing people from the excavation face and other risk areas will eliminate many safety issues, and the resource savings can be reallocated to other parts of the project.

Some fear that automation and AI will make tunnel design and construction less dependent on people and so result in job losses. But it is important to remember that the demand for safe and affordable tunnels far exceeds the current supply.

Automation and AI will allow tunnelling companies to take on more work, creating hundreds of jobs per tunnel project which previously would not have been justifiable on cost grounds.

Also, faster mining demands faster delivery of materials and testing etc, so more jobs will be needed to meet the demand at manufacturers’ and subcontractors’ sites. On-site mentoring and input from automation and IT experts will also be required so it is very likely that AI will be an important job creator in future tunnelling scenarios.

Heavy civil engineering has traditionally been slow to adapt to new technology but we must strive to do so at the earliest opportunity. No project manager wants to save an hour by using fully-automated equipment only to waste hours debugging and fixing programming/coding issues.

Current machine learning, self-improving algorithms, virtual reality (VR) and augmented reality (AR) are transforming tunnel projects globally with the help of AI and automation. AIpowered robots will revolutionise many construction tasks while engineers and project managers can use devices to simulate training, view construction progress, impose alternative scenarios in a real setting, and better visualise the end result.

Some examples of new and notable technologies are:

  • One-stop-shop modelling, simulation and analysis software that includes all aspects of design related to tunnel projects (geotechnical, structural, hydrogeology, thermodynamic, fluid dynamics, etc.).
  • Use of drones for site inspection, aerial survey, automated mapping and photogrammetry.
  • Advances in safety, including but not limited to detecting operator discomfort, sleepiness etc via cameras and face detection technology; managing/controlling safety procedures (tag in, tag out by fingerprint or eye scan); detecting human movement and energised lines around the equipment.
  • Automatic alignment control and ring selection systems for TBMs that calculate and determine the best sequence and orientation of the rings.
  • Autonomous jumbo drills and shotcrete robots programmed and optimised to perform the excavation cycle with no human operator.
  • Augmented reality and video game technology for mapping existing utilities and trenchless design of new services.

However, a few uncertainties must be evaluated, including:

  • To what extent we trust AI for the safety of humans?
  • What are the risks when AI malfunctions or crashes?
  • Can we protect these systems from hackers / cyber-attacks?
  • Can local and global regulations keep up with issues related to the use and misuse of technology?
  • How does automation change the job market for consultants and contractors?

While the benefits are clearly numerous, the challenges we will face in the next few years should not be overlooked. Safety in all forms – whether personal, property or intellectual data – must always be key.