Blog

Construction and engineering face a well-recognised skills challenge. Talked about for many years, the search for solutions tends towards attracting and developing new talent – something that must continue. But our industry also has a notorious productivity deficit.

Without significant productivity gains, the industry is unlikely to close the skills gap. Worse still, the growing complexity of project delivery likely means that gap will worsen if we don’t work out how to help our people deliver more.

In manufacturing, the solution to this has been known for generations. From the invention of the linear production line to widespread adoption of robotics, productivity has been central to that sector. And now, finally, we are seeing the very earliest signs of robotics opportunities on construction sites.

This leads us to Northumberland Street in Newcastle, where the Taylor Woodrow Civils North team were faced with electrical cables encased in concrete beneath the ground where the plans specified that drilling was needed. Rather than default to conventional methods, Taylor Woodrow saw the opportunity to trial Baubot’s advanced mobile robotic system, which is capable of drilling large-diameter, deep, angled holes in a precise pattern.

Our industry’s knowledge base for something like robotics safety is too small – and working with specialists has taught us a lot in a very short timeframe

Taylor Woodrow engineering manager Daniel Ferneyhough explains: “This is very public place, we have live cables beneath the surface that aren’t carrying current and so aren’t picked up by the usual surveying. That represents a safety risk, so there’s a big desire to move people out of the breaking ground interface if possible.

“There are massive health and safety benefits to doing that because a lot of service strikes are with people breaking ground, and that interface is far from ideal for their safety. It can also be a slow process to do this work by hand drilling, so an autonomous hole-driller is safer, as well as being efficient and quick.”

Quick understates the potential revealed by the trial. Vinci Construction head of innovation and transformation Philip Reid says, “at one point we had it drilling 10 holes in 10 minutes, accurately and safely”.

But while speed is good, it is no accident that safety is at the forefront of minds with this trial. Taylor Woodrow started working with Baubot in Austria the previous year and has pulled in safety specialists to ensure that workers and the public are safe when using the robot.

“If it’s deployed on a site with 30m of space around it, you can have a person sitting away in a deckchair to monitor it,” says Reid. “But in close proximity to the public on a site like Northumberland Street, the controller stays close by with a dead man’s handle in case it needs stopping fast.”

Taylor Woodrow contract manager for Civils North Mark Lindsley explains: “We brought in external advice and that was critical for us because our industry’s knowledge base for something like robotics safety is too small – and working with specialists has taught us a lot in a very short timeframe.”

That led to a very open process within the business as the trial team set about making the project team on site feel completely comfortable about trialling the machine. “Early engagement helped with this,” continues Lindsley. “We knew about the trial six months ahead of time, so we sat with the team, and they asked 30 or 40 questions. Even with some duplicates that had already been considered, there were questions we hadn’t had before. That was invaluable as we could raise them and ensure we had something in place.”

It also led to a positive engagement on site. “We expected some pushback from the team over this, but actually, people got really engaged with it and intrigued about what it could do.”

What the robot did was impressive. Along with reaching a speed of 10 holes in 10 minutes, it was given a programmed task plan to follow automatically, and drilled each hole to its specified depth, diameter and angle in precise configurations – in this instance to depths of 400mm and diameters of 30mm.

That said, part of the success of the trial was about identifying challenges as much as proving prowess. Reid explains: “It’s about trying things for more than a day. You have to reset what you consider failure and success to be. In our industry, we build a thing, and that’s success. Or we make a better place within a town, and that is success. But with innovation it’s different.

“Yes, there’s the success of the drilled holes, but there’s also the success of what we learned. There’s improved familiarisation and things we now know that don’t quite work yet, and we know a bit more about what potential there is to improve it and do things differently with it. All of that is success too, when looking at a new technology like this.”

This reflects the early-stage nature of the technology’s adoption. Its regulatory clearance was only achieved in April and the team involved was virtually starting from scratch when trying to understand how to ensure the trial could go ahead.

Ferneyhough explains: “We went out to Austria to meet the robotics team and even before that we had a lot of conversations about health and safety requirements. For example, we even had to establish things like who signs the permit to penetrate. We held a lot of internal workshops where people threw lots of concerns onto a piece of paper and then met with Baubot to discuss them.”

One important aspect of this is that while the machines can follow a set programme to deliver work autonomously, they do have a controller who has the power to start and stop the machine. That controller is also the person who signs the permit to penetrate, and is in charge of the equipment and its activity. Behind all of that is another human being who programmed the pathway to identify where to drill and map out the machine’s work.

Along with reaching a speed of 10 holes in 10 minutes, the robot was given a programmed task plan to follow automatically, and drilled each hole to its specified depth, diameter and angle in precise configurations

Not everything works first time

As with any trial, challenges emerge throughout, and the team was as interested in what didn’t work as much as what did.

“The team came into this expecting a certain area and were expecting to be drilling though deep concrete,” explains Reid. “But operational issues meant we had a slightly different area, we came across connectors for power supplies and cables, we had heavier rainfall than expected and some materials we were drilling into weren’t as expected.”

Lindsley adds: “One of the big issues on site is that it’s a historic road. There’s been a lot of concreting over the years, there’s cable not carrying current – but it is live and cannot be picked up by our devices. So, it’s not like running a trial on a factory floor.”

Ferneyhough explains that this slowed things down with the trial. “We have to be very dynamic on site, and at least 50% of that is dealing with challenges and changes, rather than doing what the planned day job would be. So, bringing this on site, the electric cable being found took time for the guys in Austria to reprogramme into the robot’s route and activity. Likewise, we found that concrete was less deep in places, so we didn’t have to drill as deep. All that leads to a workflow aspect as planning then went into the next hole.

“Independent judgement on site is perhaps easier, while the Baubot trial required more people involved in doing that.”

While the trial itself ended up taking longer than the work might have done by conventional methods, the team saw enough in the work it had done to move forward.

“The first one is always slower. We probably have another trial to go, to get to where we want to be, and we have to bring the clients with us as well, so that within a year we think this could be deployed on a business-as-usual basis. But there’s only around a dozen of these machines in the world right now, so scaling is an unknown quantity, and that has implications for costs.”

Cost and innovation space

One thing NCE hears a lot from industry is that innovation is a challenge because it can be hard to create the space to innovate where there are tight margins, risk-averse clients and a strong focus on safety first.

This is why Vinci Construction appointed Reid and his team to help move the company forward through new innovations. And they haven’t stopped there.

“Our senior directors and the CEO deserve credit for creating my role and my team to manage innovation,” says Reid. “Our team has created that space for innovation structurally within the business.”

Ferneyhough points out that it goes further than that. “It’s in the business plan now that each business unit manager has to have one innovation each year – so each MD has to sponsor at least one transformational innovation.”

That may not always be easy with clients being concerned about risk, but the Taylor Woodrow team take the view that this is their own responsibility.

“This trial was contractor-led. We said: ‘We are going to do this’, rather than ask ‘Can we do this?’,” says Reid. “We can understand a client becoming a blocker because they are uncomfortable and aren’t expert in this, but in 10 years this may be normal, because we need automation – so we pushed this as a contractor-led project.”

Lindsley is clear about why this has to be done. “Utilisation is central to productivity and where we’ve got repetitive tasks and big jobs where this kind of machine can be deployed, isolated and re-used, there’s a real case for this to drive efficiency. On Northumberland Street we only had so many tree pits to dig, but there are opportunities where there’s a lot more space and a bigger task to use it multiple times to do one thing again and again and again.”

And while this trial focused on drilling, the tasks the Baubot robot could help to automate may rapidly grow in number.

Reid explains: “Robotics companies are picking a first use case, creating a single use robot. But this equipment could place slabs or do other things simply by changing the front installation from a drill to some other tool. You have a track machine with a robot arm – so you could use it to break things with a drilling arm, but if the technologists develop a gripper for it, it could do block paving too.”

This could have big implications for productivity. “At the moment, the robot needs one person watching each machine,” says Lindsley. “But long-term we could see four people watching 20 machines doing multiple tasks across a single site.”

Like what you’ve read? To receive New Civil Engineer’s daily and weekly newsletters click here.