The open dome looks like a foam igloo rising from the asphalt. It's a design that would have been tough for construction workers to build quickly using traditional methods.
But it's not a problem for the DCP — a Digital Construction Platform built by researchers at MIT.
Think of it as a souped-up self-driving 3D printer on wheels. It built the scaffolding of this unfinished structure in just 13.5 hours, and all it's waiting for now is concrete to be poured.
While robotics and automation have revolutionized the way factories build cars and computers — helping companies build new types of products with speed and efficiency not possible or practical with humans alone — industry experts admit that technological advancements in the construction industry have been relatively slow.
"Construction sites are very different from the majority of workplaces in that most of the work takes place outside, in highly unstructured environments," says Dan Kara, director of robotics research at ABI Research — the opposite of an Amazon warehouse or Tesla assembly plant, say.
And so for the moment, the market for construction robotics is still in its infancy — too small a sliver to separate out of the $22.6 billion Cdn the industrial robotics sector was worth in 2016, Kara says. But the DCP is the latest in a string of recent efforts that are expected to one day transform the way that homes, offices and other structures are built, starting with this open-air formwork dome.
Safer, cheaper, faster — and even more creative
The Swiss Federal Institute of Technology in Zurich unveiled a bricklaying robot in 2015, while the Eindhoven University of Technology showed off a 3D concrete printer the same year. At DARPA's Robotics Challenge, humanoid robots used power tools to complete (or, rather, attempt to complete) certain tasks.
And slowly, there are commercial companies entering the space. Blueprint Robotics in New Jersey has a facility where home-making robots build prefabricated walls, roofs and floors. In New York, Construction Robotics' Semi-Automated Mason (SAM) does bricklaying, too. Another company, WinSun, has been 3D printing concrete houses in China. MIT has been working on the DCP and systems like it since 2011.
The goal is to make construction safer, cheaper, faster and even more creative, says Carl T. Haas, a Canada research chair and University of Waterloo professor specializing in construction and automation. "You tend to automate the stuff that's boring or physically dangerous, and hopefully what's left is the really interesting part," he says.
And with the global construction industry expected to grow from $8.8 trillion US last year to $10.1 trillion in 2021, according to Construction Intelligence Center (CIC) data cited by Kara, the coming years could be see a boom in new tech. The CIC says Canada is expected to spend $87 billion Cdn on construction this year alone.
MIT's DCP consists of a robotic arm that can 3D print structures made of insulating foam at architectural scale — in other words, big enough to live in. In a test in July 2016 the team printed an open dome 14.6 metres in diameter and 3.7 metres tall over a two-day period, and said they could have gone higher.
"We believe that, at the time of writing, this printed test structure is currently both the largest monolithic structure ever 3D-printed by an on-site mobile platform and the fastest autonomously printed architectural-scale structure," the researchers wrote in a paper published in the journal Science Robotics this week.
Haas says 3D printing in construction is one area "which is absolutely going to explode" — but there are differing views on what works best. Do you prefabricate structures or print them directly on-site? Should the robot be mobile or fixed in one place like a gantry? And if you choose mobile, should the robot drive around on wheels or take to the air like a drone?
- VIDEO | Dancing humanoid robot rolls in to Calgary banks
It's a challenge made tricker by the fact that in construction, "every building is bespoke, in a sense," says Haas. There are variations in buildings codes, customers needs, site conditions, supply chains, materials and weather. There's not one type of bot that can do it all, and the industry — fragmented as it is — hasn't yet taken in a big way to any of the few products that do exist.
'It's going to be a huge wave'
Still, there's more automation in the industry already than you might think. Excavation and grading is essentially robotic now, Haas says, driven off 3D designs. Self-driving vehicles are already used in big construction projects like mining. Computer vision has also "exploded," and is used to automate measurements for custom jobs.
One area where robotics are already having great impact, Kara says, is prefab construction, an industry worth $38.3 million Cdn in 2016.
The MIT researchers say future iterations of autonomous machines could build structures in extreme or inhospitable environments — say, the aftermath of an earthquake, or even on another planet. There's also an opportunity to collect new types of data on how buildings are constructed (think: quantified construction). Optimism abounds.
Haas has been in the industry for 30 years — he's vice-president of the International Association for Automation and Robotics in Construction — and compares construction robotics research to work on AI, both hyped in the 1980s before a decades-long lull.
"It's kind of in a renaissance right now," says Haas. "And I think it's going to be a huge wave this time."