Data centres feature standardised designs and their construction favours modularity. So surely they are ripe for the adoption of robotics and automation? The construction engineer from one of the world’s largest data centre developers has lifted the lid on this booming sector.
Whichever data centre developer list you review, Digital Realty is always in the top 10 – and often at number one. Stephen Henley, vice-president, design engineering and construction, EMEA, at Digital Realty, told the audience at this week’s GroundWorks event in London about the challenges of introducing innovation on such high-value, high-risk projects. Digital Realty currently has between 50 and 60 data centres being built in EMEA and 30 in APAC.
“We tried the robotic drilling rigs, I think it was the Hilti one, in Barcelona. The problem we have with them is the arms aren’t long enough. So we need ones with longer arms, please,” he said. “The hall space is about 8m to 9m tall, and I think the drilling rig can only go to about 6m or something like that. So we had to put it on a platform, and it starts to lose its efficiency.
“In the US, because of the scale, we do a lot of the setting out tools, the scanning tools. We’re not seeing so much of the actual robotics for helping construction. I’d like to see how the market’s going to address that.”
New technology, new workflows
Henley and Digital Realty are interested in “anything that increases productivity, increases output, reduces rework”. He accepts that return on investment (ROI) from innovations may not be immediate, but the potential return must be visible.
“It’s not necessarily about an immediate ROI, because you’ve got to be realistic. If you’re bringing a new product, a new machine, technique or workflow into a space, it won’t be that efficient initially. But you’ve got to be able to see through it and what it could do in the future.”
He noted the difficulties bringing new technology or processes into an existing process: “You’re changing the whole workflow of that particular sequence of events, and that takes some time to do, because it’s okay for me to tell the team that they’re going to use that, but then they’ve got to get to the designers, get to the contractors, get to their subcontractors – and they all have to believe in it as well, otherwise it doesn’t get the traction it needs.
“We’ve been through some big cycles in the last two to three years to [adopt] a lot of these preliminary software products for productivity, but it’s taken a while to get them embedded. But they are now starting to make a difference, and predictability scores are going up significantly. It’d be the same for any robotics.”
Top-down approach
Digital Realty favours a top-down approach where it specifies products and systems to its contractors. “If we’re looking at robotics, it will end up being us that will supply it and then novate it onto a contractor,” Henley declared. Asked whether this approach creates friction with contractors, he answered: “It did initially, but now they’ve come to realise their productivity is going up, so they’re more predictable, and the days beyond due are now fewer. We do a lot of tracking metrics: we’ve halved our days beyond due.”
Henley said Digital Realty is prepared to tweak its designs to enable the greater use of robotics. He explained the developer’s process: “When we’re looking at the design of a data centre, we’re designing it for speed. One of the key scope requirements for our engineers and architects is to design something that can be assembled very quickly.
“We run it on what we call the production control model. There are production labs on every site [a mission control room – ed.], and we go through all of the productivity output. We look at what’s being planned over the next eight-, four-, and two-week look-aheads, and look at how safety is being monitored within those periods as well.
“We look at all the high-risk activities, and the more we can start to bring automation into that, the more I believe we can become more productive. But to deploy robotics into a fast-track production line, I’m not sure we’re quite there yet.”
ROI: double or quits
The most effective innovations are those that deliver more than one benefit. So said HS2 senior innovation manager Ben Wagenaar, who spoke alongside Digital Realty’s Henley.
Reflecting on return on investment (ROI) and business cases, he said: “The best-performing innovations in all areas, not just robotics, have been where they’ve had two or more big wins, big creative leaps. [If not] the juice is not worth the squeeze. If it’s just one thing that gets you faster, you’re into that [apocryphal] Henry Ford quote, ‘if I’d have asked my customers what they’d wanted, they’d have just said faster horses’.”
HS2 is committed to innovation, Wagenaar said, whether through the contracts with the joint venture delivery partners that require the latter to bring innovations to the project, or through HS2 itself. “We do care about how things are delivered. Our joint ventures, in their contracts, have to provide innovation services to us.
“Our contracts work both ways really. There’s no stipulation of what technology to use, but similarly, we’ve encountered situations with our supply chain where we’ve said, ‘we’ve got some funding here, we want to try something different’. And they said, ‘it’s not in the contract, can you help us, please?’
“In nearly all cases, they’ll come with us because if it’s trying something new and it’s low risk, they might learn something. If there’s something that they want to bring to our programme that’s going to help, we’re all ears and we’ve got a funding route for that.”
