Imagine a piece of software that could help you choose the route of a new metro crossing a city as complex and constrained as London.
One that would tell you exactly which route to select to weigh the competing constraints, and find anoptimum balance of construction and in-service operational efficiency.
Consider another bit of technology that could operate remotely underground, and pinpoint where there may be defects, directing inspection staff straight to locations of concern.
This could be the future of developing new underground infrastructure, and maintaining assets in an operational environment with limited access – and it is not that far away, according to engineering consultancy Atkins.
3d data 2
The firm believes automated design and artificial intelligence could dramatically change the way civil engineers operate. Technology would allow better utilisation of engineering skills, bringing new insight and efficiencies to the development and maintenance of major tunnel infrastructure.
We cannot continue to work in the same ways, says Atkins head of tunnelling Ian Gee; “We are in a period of rapid technological advancement.
“There are a lot of social and political change ahead of us, but as a practice we want to focus on the opportunities from technological change.”
The role of the engineer will change.
Atkins head of tunnelling Ian Gee
Atkins is optimistic about the opportunities from technology and is empowering its engineers to explore them.
“We are moving from piecemeal application of digital technology to having it front and centre of our business,” says Gee.
“Our engineers are hugely enthusiastic about exploiting technology; we’ve asked them to do some blue sky thinking and to try things out.”
Point Cloud Survey Used in Conjunction with Auto Correction Algorithm
One such innovation that Atkins has trialled is computerised image recognition, using artificial intelligence (AI). Technical director Christopher Chew says this represents a major advance in the use of technology to process visual data.
“Computing algorithms normally deal with numbers rather than pictures, so this is a step forward, as cameras, data storage and computingbecome better, cheaper and faster.
“The industry regularly collects images of assets using vehicle mounted cameras. The use of drones in areas that are difficult to access is advancing. If we can process visual information effectively and repeatedly, we can improve maintenance efficiency and prevent defects affecting network operations, providing valuable long term benefits.”
Today powerful free software can be found online, and there is no shortage of data
Atkins technical director Christopher Chew
“In metros, the periods available for maintenance are shortening with competing pressures to operate 24 hour services while caring for the ageing infrastructure,” says Chew.
“Technology can greatly assist the human inspector, thus improving network safety and reliability”.
The process, based on AI code, has been trialled by Atkins in the water sector. Gee says. “We are working with a partner company using AI to recognise other defects and other features captured in digital photographs.
point cloud surveys
“We have used it on new build water tunnels in the North West to recognise and remotely log features. We are extending that and have several clients very interested in the technology, including the European research body Cern.”
He adds that while the current challenge in the industry is to develop programmes and processes capable of finding known isolated physical defects consistently from photographs, the future of the technology is to compare and contrast “changes in conditions” over time.
We will need that diversity in the future more than ever to help drive innovation.
Atkins head of tunnelling Ian Gee
“A change in leakage pattern for example can be the trigger for a human eye to take a closer look, using the technology as a rapid screening process.”
Another ambitious area of technological advance Atkins is pursuing is the ability to create programmes to optimise complex tunnel infrastructure designs.
“Modern engineering is about speed,” says Gee. “We now can do in a morning what used to take months.”
“Our brightest digital creators are looking at using Autodesk Dynamo Studio to try to automate the design process to do iterations very quickly. We are trying to apply increasingly complex, rule-based engineering.”
This could mean computers effectively making complex design interface decisions by applying a vast number of ordered and finely tuned rules, defined by engineers.
Chew says this technology would help with fine-tuning route decisions on complex schemes.
“On high speed rail, if you alter the rail alignment within a tunnel section, you can quite easily impact the adjacent contracts,” he says. “Rule based automated design would mean you could input your changes and the technology would ripple through and identify all the potential impacts elsewhere, allowing engineers a complete overview of their decisions.
“When technology includes a hierarchy of rules, it becomes very robust. This could significantly improve clarity and decision making during the planning phase for major schemes,” Chew believes.
Again the technology is available and is being tested on small projects.
“We will look to expand it,” says Gee. “Crossrail 2 for example, is as complicated as it gets. The sponsor recognises the value of centrally coordinating key planning information. The development and implementation of a total digital route selection system is ambitious, but it is the future of total digital engineering.
“As the industry prepares for Crossrail 2, technology continues to rapidly advance and there’s an expectation that things will need to be done differently. Milestone major projects such as the Channel Tunnel, the Jubilee Line, High Speed 1 and Crossrail have acted as technology catapults, and I can only see this intensifying in the future,” adds Gee.
“The role of the engineer will change,” predicts Gee. “People will need some different skills to build and run the technology.”
He remains confident about the role and ability of the engineer, adding: “Tunnelling and ground engineering are knowledge based industries; it’s difficult to commoditise the work within them, so whereas the way engineering projects are delivered may change, the essential, experienced-based judgement will remain important.”
He believes the major challenge for the tunnelling industry – and for many other sectors in the modern era – is to find ways of using data and technology effectively.
Chew believes data quality remains a key ingredient of technology.
“Today powerful free software can be found online, and there is no shortage of data, but it’s all about organising and putting data sets together in the best way to squeeze value from them,” he says.
Having keen minds in engineering will remain critical, both to drive the innovation and to use it.
“We have a diverse workforce,” says Gee. “More than 20% of our capability is women and rising, and we have a lot of non-UK nationals.
“We will need that diversity in the future more than ever to help drive innovation. ”