How bots will build tunnels of the future

HyperTunnel joint chief executive Steve Jordan is not a man to mince his words. “Current tunnelling technology has reached its physical limits,” he insists. “It’s too expensive, too slow, and still too risky. You can’t turbocharge a TBM [tunnel boring…

HyperTunnel joint chief executive Steve Jordan is not a man to mince his words. “Current tunnelling technology has reached its physical limits,” he insists.

“It’s too expensive, too slow, and still too risky. You can’t turbocharge a TBM [tunnel boring machine]. Something radical is needed.”

That radical new approach has already secured the firm a contract with Network Rail too.

Expertise from many sources “Our new approach combines expertise and technology from many sources, from opencast mining to 3D printing, using virtual reality (VR), BIM [building information modelling], digital twinning, digital swarming and artificial intelligence (AI),” explains Jordan. Jordan says he has no background in tunnelling, just general engineering experience with an emphasis on production engineering. Frustrated by the options on offer for a tidal range project which involved tunnelling, he assembled a think tank of experts from diverse fields – including aeronautics and Formula 1 – to develop a new tunnelling concept. He deliberately omitted anyone from the tunnelling sector. “We didn’t want to be encumbered,” he says. What emerged from their deliberations was truly radical blue-sky thinking. Move beyond the technology of the last 150 years, keep human operatives safely above ground, and let the drones take the strain. Semi-autonomous drones are commonplace now. Drone swarming, in which hundreds, even thousands of such drones co-operate to carry out specific tasks with the minimum of human intervention are a realistic option. Swarms of drones – or underground “bots” – is the technology at the heart of the HyperTunnel approach, key elements of which were due to be unveiled at the British Tunnelling Society’s (BTS’) conference and exhibition at the QEII Conference Centre in London shortly after this issue of NCE went to press. Our new approach combines expertise and technology from many sources, from opencast mining to 3D printing A typical HyperTunnel project would begin with the insertion of an index bore, a 225mm internal diameter high density polyethylene (HDPE) pipe, through the centre of the proposed tunnel, using coring technology from the oil and gas industry. Core samples yield the first data on the ground conditions. More index bores might be installed. These then become “botways”, into which can be inserted robots – or “bots” – carrying a rotating ground penetrating radar or seismic sensors. “HDPE is a very useful material,” Jordan reports. “It’s strong and waterproof, but it doesn’t block radiation [from ground penetrating radar] and it’s easy to drill into.” A standard HyperTunnel bot is relatively simple. Each single tool section is less than a metre long. It has a semi-circular cross section, is wheel mounted and battery driven, with a central compartment into which a range of tools can be installed. One of these is the HyperBit – a metal tipped, extendable plastic nozzle with a one way valve that can be drilled out through the pipe wall to allow access to the local geology. Dozens of these HyperBits can be installed by a single bot. Sensors deployed through these access points can collect and feed back data on soil moisture content, chemical composition and temperature. More botways are then installed and more data collated. Eventually, it is anticipated, there will be enough data to construct an accurate virtual reality model of the proposed tunnel’s geology and identify any areas that may pose particular problems. “An individual bot can only work on a small scale,” Jordan explains. “But if you have thousands of bots working simultaneously, then progress is extremely rapid. And these tools are working to an accuracy of tenths of a millimetre.” We could install a new lining outside the existing lining while trains continue to run This implies dozens, maybe hundreds of bots working in a single botway, all needing to shuttle to and fro to load, unload or recharge. Coming up with a design that allows bot to pass bot was one of the biggest technical challenges, Jordan reports. “There will be bots heading out, bots heading in, and bots in position undertaking tasks. Some bots will have priority over others. Yet they have to be able to pass each other at reasonably high speed.” A key part of the solution was the semi-circular cross section, which meant there is just room for two bots to squeeze past each other back to back. The only practical way of achieving this is for each bot to turn on its side, to which end every bot is fitted with turning vanes front and rear. To prevent a bot toppling over as it rotates onto its side, an anti-roll bar extends. This then has to be retracted at the last possible microsecond as the two bots begin the passing manoeuvre. Working with Network Rail This summer the company has been working with Network Rail, which Jordan reports is particularly interested in HyperTunnel’s repair and enlargement potential. For Network Rail, HyperTunnel will be tasked with drawing up radical cost and delivery time improvements. Products being trialled include digital twins, robotics, 3D printing and digital underground surveying, supported by AI and VR. “There are a number of old tunnels that are too small for modern freight trains, too restrictive to allow for electrification,” Jordan says. “We could install a new lining outside the existing lining while trains continue to run. This would mean a much shorter closure, just to remove the old lining.” Although Jordan insists that the basic technique would work even in hard rock, the company is concentrating on projects where the relevant geology is London Clay, Chalk or gravel. “That’s 75% of the market,” he points out, “And we’re a relatively small company. Yes, we could build a tunnel on the seabed, but that’s for another day.” There is another significant advantage over conventional methods, he adds. “There’ll be far less surface disruption, a key factor when working close to sensitive areas. It makes it possible to consider tunnelling under ancient woodland, for example, rather than excavating a cutting.” Despite turning an initial cold shoulder to the tunnelling community, Jordan and his business partner Jeremy Hammond have now assembled a heavyweight technical board, with some of the biggest names in the sector – including Vinci Construction Grands Projets managing director François Pogu and Peter O’Riordan, formerly Crossrail 2 package manager – throwing their weight behind the HyperTunnel alternative. HyperTunnel has now set out its stall for the wider industry to see what could well be the future of tunnelling in the 21st century. The official unveiling could be the seminal moment. Like what you’ve read? To receive New Civil Engineer’s daily and weekly newsletters click here.