Insights and best practices for construction management technology and project delivery

Walbridge Director of Innovation on the Future of Machine Learning

In Part I of our interview with John Jurewicz, director of technology innovation for Walbridge Building Design and Construction, we explored John’s professional background as well as his insights into emerging technologies.

In Part II, Dr. Burcin Kaplanoglu, Executive Director, Innovation Officer for Oracle Construction and Engineering, and John discuss the potential of machine learning and where technology is heading long term.

BK: What is the biggest value in AI and machine learning leveraging the existing data we’re collecting in the short and mid-term?

JJ: The value of machine learning—or algorithms that optimize—is going to first evolve in repetitive building types. For example, you’re putting up data centers and modular construction in days instead of months.

You have a pattern of what we've done in the past, and now we want to increase efficiency or look at how to shave 30 percent off the schedule, which obviously saves money. The initial result you see is we’re taking the patterns that you’ve already collected in your data and optimizing it with better intelligence.

It’s usually experience-driven—people looking at the algorithms but also saying, “OK, if you're checker boarding the concrete pours to optimize how the concrete cures, what if you use two crews versus three?” Study how the crews affect each other’s efficiency.

Monitoring patterns for optimization

For example, we can see the weather cycles and the predictability of rain. What if we build at a different time of year, when we're not as likely to encounter rain or flooding? You’re going to see a mixture of weather data with predictability of patterns of how you pour concrete.

And you see some of these technologies we’ve been talking about. We’ve seen algorithms with recipes that can improve how you self-perform an activity.

But I’m thinking more holistically about the following optimization questions:

  • Should you even be building like this?
  • Or should you be building it at a different time of the year, and if so, how much money would that save?
  • How can you optimize by evaluating four different sites where you could potentially build?

By observing the patterns of how to deliver concrete or build a batch plan, you can optimize the site selection based on the construction value or by lessening the environmental impact.

You’re weighing both considerations simultaneously: “I’ve got less erosion and less impact to the environment, but am I paying a premium to deliver concrete or services to this site?” You can begin evaluating things differently.

BK: Where do you see technology heading in the long term?

JJ: I foresee two trends in our industry:

  1. Constructors will become more technological in their approach to services.
  2. We’ll also be more driven towards prefabrication. We’ll partner with people who are very smart. Or we will become very smart at building portions of buildings, meaning we’ll specialize in the prefabrication of certain building types.

For us on the industrial side, if we’re designing and building data centers near power stations that are cogeneration facilities, we may start specializing in building just those because the cost of distributing the electricity is so much less.

We’ll become advisors on how to build buildings for the biggest bang for the buck, meaning you build these factories close to where power’s generated. And that isn’t new.

If you build something near a hydroelectric dam, you could say it’s going to be a lot easier to get reliable energy. But I’m thinking more in terms of looking at your existing nuclear assets.

The cost of energy is going to go down. They’ve got to do something with these nuclear facilities which are very expensive to run. Why aren’t they locating the data facilities near the power stations?

Or, why aren’t energy-intensive factories that are producing gypsum drywall located closer to where the power source is?

Tracking emerging projects will become easier.

Expensive tools are rapidly becoming more affordable everywhere. It’s going to almost be an expectation now by owners: "Well, why aren't you tracking and doing things that quickly? Because everybody else is.” 

Owners are beginning to demand tracking work in place, using more meaningful dashboards, and telling you what's going wrong in terms of risk sooner. You can track bad news much more quickly.

For example, Intel's RealSense sensors, which are under $200, are essentially doing what the $500 Tango did last year. And for quick little scanning studies, why not? 

It’s better to track work put into place. And once you know exactly what’s put into place, what do you do with it? How do you adjust and start to optimize or reduce people working on top of each other - like you see with Oracle Prime Projects?

We can pull people apart so they're not working on top of each other. That's reducing our risk. There's a real value there. 

Those construction advisors are going to evolve long-term. We’re going to become better at giving good guidance on where to build in the future.  And you’ll be able to back it up with data using AI for sure.

Discover more innovative ideas emerging in construction and engineering in our “Navigating the Future of Projects” report from Oracle Industry Connect.

Related posts:


Be the first to comment

Comments ( 0 )
Please enter your name.Please provide a valid email address.Please enter a comment.CAPTCHA challenge response provided was incorrect. Please try again.