The Future of Autonomous Construction Vehicles

AI on top of more and more sensors will take us to places we’ve never been.

Autonomous construction vehicles are already at work on some jobsites, and more are coming soon. These machine are performing or will perform relatively simple tasks such as materials moving, pile driving and excavation. But more complex tasks demand more complex machines, and developing them will require advances in technology.

Non-repetitive tasks are the Achilles heel of construction, often blamed for stagnant productivity in the industry. They are not easy to automate.

“That’s always been the issue when you talk about automation of jobsites; every situation is just different enough that it is really hard to create any kind of standardization and automation,” said Helge Jacobsen, United Rentals’ vice president for operations excellence and general manager of the Advanced Solutions Group, which partners with original equipment manufacturers to develop and implement autonomous and IoT-enabled solutions for customers.

“The answer is that we will have more and more sensors and artificial intelligence built on top of those sensors, so the equipment gets smarter and smarter to adapt to all these different situations,” Jacobsen explained.

More data equals smarter machines

“The way artificial intelligence works is that you need to feed it data. It becomes a chicken and egg issue, where you need to have enough applications out there that you can collect data the computers can learn from, and then come up with new solutions so they can go out and get even better.”

That data collection is already happening. “More and more of us — United Rentals is one, we are collecting data from the entire fleet on a day-to-day basis — are starting to put sensors on to capture the data. As you start getting these data sets, you have enough data for the artificial intelligence engines to work on.” Once those engines learn from that data, new solutions will evolve. And once those solutions hit the field, “now we can collect even more data, and then it’s suddenly going to go really fast.”

Data collection is the reason Facebook can recognize cats, Jacobsen noted. The millions of pictures of cats that have been posted gave Facebook a massive data set to work on. “Now they can suddenly identify cats and they can get incredibly precise about the cat — the type of the cat, the mood of the cat, all of these other things.” That’s the power of data.

Choosing the right tasks to automate

Choosing the right tasks to automate — creating autonomous solutions that will yield the most value — will be key. Materials handling, a huge source of time waste, is a perfect example of the right kind of task to automate.

An example of the wrong kind of task, according to Jacobsen, is apple picking. There’s a machine that can identify the shape and color and firmness of an apple and pluck ripe apples off the tree — “an absolute marvel from technology standpoint. ” But humans already excel at identifying ripe apples and can pick one in half a second. A better task to automate, Jacobsen noted, would be walking to and from the truck with 40 pounds on apples on your back.

“When we talk about autonomous equipment with our customers, what really resonates with them is looking at the value stream, identifying where the biggest portion of waste is and then trying to find solutions for that,” said Jacobsen. 

Will automated equipment replace workers?

Even as automated equipment evolves, the need for skilled workers, especially craft workers, will remain. For example, SAM, the semi-automated bricklaying robot, can lay bricks quickly and accurately — a boon for contractors, since bricklayers are in short supply — but a mason is still needed to build up corners and load the bricks for the machine.

Automated equipment will free up time for skilled workers to do more critical tasks. With automated earth movers, for example, excavation can happen on weekends and off hours, allowing the worker to apply craft and skill where it adds the most value.

With AI learning driving the automated solutions of the future, it’s impossible to say now what those solutions might be. But one thing is for sure: As autonomous equipment becomes more prevalent on jobsites, it will transform the construction industry. “We will start changing our processes based on the new technology, and the new technology will enable new processes,” noted Jacobsen.

“Ten to 20 years from now, our jobsites will look dramatically different than they do today.”

 

Marianne Wait is an editor and writer who creates content for Fortune 500 brands.

Image Credit: Zhu Defing / Shutterstock.com