“I’m an old-timer who likes working with my hands. Craft was always an important part of what I did–depending on my back, my strength, and my hands to provide for my family, and hope for better things for them.”

That’s Rich Holbrook, Director of Construction Operations at Layton Construction Co., a nationally ranked, 64-year-old commercial contracting company with more than 800 employees.

Holbrook cherishes the old ways of doing construction, but also gladly embraces new ways. He believes that today, in order to stay relevant, every GC must embrace construction data and machine learning.

Demystifying Construction Data and Machine Learning

The words “construction data” and “machine learning” cause a lot of eyes to glaze over. Holbrook admits the concept is intimidating to a lot of people, and that’s why some hesitate to implement it. But it’s not as complicated as it sounds.

Construction data simply refers to the documentation and information that every construction site already accumulates. For some contractors, this documentation takes the form of papers wadded up under truck seats, photos accumulating in phone storage, scribbles in a notepad, as well as all the paper trails of issues, change orders, contracts, and communications among architects, owners, contractors, and subs.

In the old way of handling construction data, this information is disconnected and often disorganized. At the end of the project, the documentation exists, but most of it is so difficult to find and sift that it’s virtually useless.

“Construction gets a lot of flak over not documenting things, but I think that’s false,” says Jeff Sample from ConTechCrew. “We’re incredible at documenting. We’re just terrible at putting it where it can be found and shared.”

Technology is helping to solve that problem by aggregating data into a central location. Platforms like BIM 360 provide a way to capture the documentation in a useful, searchable fashion.

Likewise, machine learning is just a fancy way of describing the latest innovations in making it easy to manage and utilize all that data. Think of it as a robotic tool, one that learns from data and trains itself based on what it finds. This tool sits inside project workflows, analyzing the massive amounts of construction data held there, and uses it to predict safety and quality risks. Users can access these learnings on-demand via dashboards, or investigate as-needed when the tool automatically alerts them of the day’s high risk areas, before incidents occur.

“Construction gets a lot of flak over not documenting things, but I think that’s false. We’re incredible at documenting. We’re just terrible at putting it where it can be found and shared.”

Jeff Sample, ConTechCrew

How Machine Learning Supports the Craft of Construction

“Construction used to be done with a good handshake and a trust and a faith in your subs that the work would be done and done in a trustworthy way,” says Holbrook. “But now we don’t always know who we’re working with, so we have to have measures in place to help them achieve what they’ve contracted to achieve.”

A construction management platform like BIM 360 helps in real time by connecting the data sources and standardizing it into one database, making it easily accessible and searchable. Machine learning takes it to the next level by analysing that data, and using it to help improve the predictability of the project.

“Machine learning finds signals in the data and is able to identify and predict risks,” says Manu Venugopal, Senior Project Manager for Autodesk BIM 360. “Leadership uses that information to improve trainings and to take it into their daily huddle to address issues before they become problems.”

Machine Learning Improves Safety

20% of all work-related deaths in this country come from construction, amounting to nearly 1,000 deaths each year. For each death, there are hundreds or thousands of near-misses and small incidents that never get reported.

For old-timers like Holbrook, this is an emotional subject.

“We want to make sure that everyone who steps foot on our job sites goes home in good shape,” he says, “so they can support the people who depend on them.”

Machine learning applications like Autodesk BIM 360 Project IQ help companies be proactive in identifying potential incidents before they happen.

“When good site data is captured,” explains Venugopal, “in that data we can find signals for unsafe behaviors and near misses. This is what machine learning is able to identify and predict. Then if you’re a project manager, you can drill down and see if a sub is experiencing a lot of risk, to see what’s happening. This is where machine learning is becoming part of the day to day process in the construction industry.”

“We want to make sure that everyone who steps foot on our job sites goes home in good shape, so they can support the people who depend on them.”

RICH HOLBROOK, LAYTON CONSTRUCTION CO. 

Machine Learning Improves Quality

“Project IQ takes data analytics from our project management system and puts it in a summarized fashion showing what the real risks are associated with each day, whether it’s quality or safety,” says Holbrook. “And the beauty of IQ is that I don’t have to do the filtering and searching. It’s done for me based on data gathered on all my projects.”

If there is a sub who is consistently underperforming, Project IQ identifies that and brings it to the attention of the project manager to address. Likewise, it identifies other risk factors for quality issues and makes it easy to address them before incidents occur.

“The ROI is substantial on projects that adopt this process,” says Holbrook.

Machine Learning Is Here Now and Easier Than You Think

“This is not something that is future-looking,” says Holbrook. “It’s happening today. If you’re not on something like this, you’re going to fall behind.”

Holbrook says that he’s been helped in adopting the new technologies by the fact that he has children who have joined him in the industry, and they encourage him to try new things. But the reality is that every company must begin planning for a now in which machine learning is used to manage construction data and improve the construction process.

“Take the leap of faith,” admonishes Holbrook. “Just start documenting, get a system that tracks what you’re observing, stay with it. As an older gentleman, I can say that seven years ago I’d have said, ‘get out of my face, I’m trying to build something here.’ Today I can honestly say, ‘Why didn’t I have this before?’’

“This is not something that is future-looking,” says Holbrook. “It’s happening today. If you’re not on something like this, you’re going to fall behind.”

RICH HOLBROOK, LAYTON CONSTRUCTION CO.

Want to know more? Listen to Holbrook and Venugopal chat about the present and future of machine learning in the construction industry.

 Looking to get started connecting your data & documents? 

Download the free "Beginner's Guide to Connecting Construction Data and Documents"

{{cta('0263bf39-1599-49d4-a7ff-77a8ebed6379')}}

{{cta('dc8799f3-0a2a-44f3-8084-1ae386698716')}}