Is artificial intelligence the answer to HS2’s rising cost problem?

Necessity is the mother of invention they say, and right now, HS2 needs to save cash. In response to a recent project update, which showed the high-speed rail scheme digging further into its contingency funds, the Department for Transport has told the scheme’s bosses to find savings.

In major civil engineering projects, as in most sectors of construction, calls to cut costs often translate as reductions in scope or quality, but for Jonathan Kelly, senior innovation manager at HS2 Ltd, the need to deliver savings is an incentive to find smarter ways of building.

The HS2 innovation team itself is small (in the single digits), but aims to punch above its weight by helping to develop start-ups through its innovation accelerator, run in partnership with lab space provider Bruntwood SciTech, which is jointly owned by property firm Bruntwood and Legal & General, and government agency Connected Places Catapult.

The accelerator, which recently closed applications for its fifth iteration, develops innovations for four to six months before showcasing them on “demo days” to the HS2 supply chain as well as industry investors.

In many ways, the innovation department acts as a kind of matchmaker, with the top players in a construction industry that, in Kelly’s words, “isn’t known for being on top of the game within AI”. He says the accelerator can “break down the barrier” for disruptive technology, “making sure they can have an impact with tier ones within our supply chain where they usually won’t get that chance”.

The scheme will result, ideally, in solutions that can help to deliver HS2 more quickly, more safely and for less money. The innovation department claims its work has so far saved more than £200m from phase one - but internally, they are understood to be targeting £1bn worth of savings.

Kelly says it is “really key for us that, if we’re taking someone through the accelerator, they end up on the project”. But that there are some exceptions.

“There are scenarios where it may not be suitable and they’ve been chosen because it might be impactful for the wider environment or the wider industry. We see that as a big win for us as well, if they are gaining further investment, they are growing as a company,” he says.

We’re already seeing the benefit in terms of time and cost in terms of how the civils are moving forward in certain areas - we have only just scratched the surface of that benefit

The latest iteration of the accelerator focused on artificial intelligence, seeking five candidates who can use the technology to “help smooth delivery of design changes” across HS2’s construction phase. “Cascading design revisions through a live building programme, such as innovations to reduce embedded carbon or alter any detail of its systems, stations or depots is challenging, especially one on HS2’s scale,” said Kelly.

“To help smooth the process, we are looking for innovative tech start-ups with applications to both design and game-plan implementation scenarios that will help strengthen understanding of how the best results can be achieved.”

That is not to say that AI is not already a major part of HS2 - the innovation department currently has 32 AI projects, working with more than 25 SMEs. Machine learning algorithms are already in use processing security camera feeds to monitor health and safety and ensure appropriate personal protective requirements are in use, and AI has also been employed to speed up the pace of groundworks.

“By feeding [ground sensor] data back real time into the design model, through AI and machine learning, you can predict a better model, based on the movements you are seeing. That has been talked about before on projects like Crossrail […] but this is the first project to ever have implemented that in the UK.

”It is the first project that has been able to look at the data coming in in real time, and change their models and designs quickly to understand: ‘OK, actually we don’t need to put in this temporary works because we know the ground has not moved anywhere near as much as we thought it would, so we are safe to proceed’. 

“It saves so much cost and we’re already seeing the benefit in time and cost in terms of how the civils are moving forward in certain areas - we have only just scratched the surface of that benefit.

“If that technology was there and the supply chain was convinced of that technology from the very start of the project, then we could be saving massive amounts.”

>> See also: Is the future of HS2 at risk?

According to Kelly, the scale and longevity of HS2 makes it a perfect environment for developing innovations, particularly AI-based technology. The multi-phase nature of the project allows tech such as groundworks AI to be tested and developed on phase one, to be deployed properly on the second phase.

“HS2, being the biggest project in Europe, also means we have got the biggest datasets in Europe [which] gives us the opportunity to make the most of AI,” says Kelly, adding that the project has also been able to learn from the failures of Crossrail and HS1 by pulling data from those megaprojects.

What is more, Kelly says the multi-phase aspect of the project is not only beneficial for the start-ups, but also brings new cutting-edge skills and expertise to contractors making them more competitive at home and abroad.

“By us pushing AI, [contractors] are picking it up as a parent company and using it on other projects,” he says. “With a lot of the contractors, this may be the first time they’re using [these innovations].

“But the legacy is that by learning from it they get to use it in other environments, whether it be in the UK or elsewhere […] Because they are doing it sometimes for the first time here, the knowledge is going to be invaluable for them.”