Machine Learning in Construction

Uncategorized 0 Comments

Throwing a glass of water on a robot will keep it in line. Since 75% of our planet is covered by water, we humans really shouldn’t worry about machines taking over the world. That said, the sciences of computers, linguistics and psychology are being combined to revolutionize the way computers impact construction projects from beginning to end. Leading construction companies are already embracing this change.

A plethora of data is generated for and from construction projects every day. In the old days (maybe not so long ago for some of us), human beings had to sift through the information, decide what was relevant and urgent, and issue the orders to accommodate updates as fast as they could intelligently handle them. The brain, after all, was the first computer.

Nowadays, artificial brains – computers – do essentially the same thing faster and with a much smaller margin of error. In fact, computers can basically look into the future of the project and avert mistakes before they happen.

The May, 2018 article from Connect&Construct, “What is AI and Machine Learning in Construction?” expounds upon the trend toward using artificial intelligence in the construction industry and how that is already improving safety, productivity and profits.

Some of the current uses of machine learning include predicting and mitigating risks – forseeing  potential hazards and adjusting plans in order to steer around them. In fact, ever since IBM’s Deep Blue splashed across the headlines in 1996 in its chess match against world master Garry Kasparov, artificial intelligence has steadily progressed in its ability to augment the projects that human brains begin.  

And while computer science has come a long way toward making life better for us, this is just the beginning. Described in the Building Design+Construction article, “Deep Learning + AI,” research on “Deep Learning” is seeking ways to program computers to “think” like a person – to observe, practice and find solutions that humans might not see. Another BD+C article, “How the Fourth Industrial Revolution will alter the globe’s workforce,” explains that data analysis is an emerging job type needed in order to make sense of the torrent of information besieging us in everyday life – in this case, the AEC industry. 

Computers can research data about performance of subcontractors on previous and current jobs (ie. if someone is violating safety standards on the current job) and raise a red flag with project managers. A subsequent heads-up in the daily project meeting can avoid problems later. In the way only a machine can sometimes do, computers can “stand outside” the project to see the big picture and identify safety concerns that active workers might not recognize or notice.

This is especially compelling as the project progresses, and workers become more and more at ease with the job as it’s built out. They may not see the day to day visual clues – what we refer to as the “little things” – that threaten to jeopardize safety. Lives are literally at stake in some cases, and OSHA statistics show that nearly one-quarter of all worker deaths in the U.S. each year are in construction. According to OSHA, the “fatal four” accident types that cause more than half of annual construction deaths are falls, electrocutions, being struck by objects, or being caught in/between. Any investment that can reduce those statistics is good for the construction industry as a whole.

As the Connect&Construct article points out, even the smallest construction job can have scores of requests for information, change orders, open issues, and urgent matters. An emergency will obviously rise to the top as needing attention. But what about the rest of the stuff? A computer is the best kind of assistant – one that can take all that pending data and prioritize it so the manager knows what needs attention every day and in what order. On some jobs, the key might even be as basic as taking documentation from the dashboard of the pickup truck and getting it into the computer where it can be analyzed and shared.  

The Autodesk BIM 360 platform of collaborative products is a leading option for companies looking to profit from AI and machine learning. Data sources can be connected and made consistent, accessible and searchable. Machine learning can analyze that data and improve quality, safety, scheduling, and profitability. It’s inevitable that the job processes will be enhanced.   

Prefabrication is another component in this construction revolution that promises the same project improvements. As the AGC Constructor article, “Off-Site Modular Construction Improves Quality and Safety of Projects,” points out, constructing in a controlled environment improves quality, safety and job performance. It enables more predictable scheduling – for instance, no weather delays. Construction site waste is reduced, and there is less disruption of the building site. Finally, of utmost concern in the current labor market, the process helps contractors cope with the smaller available workforce.

Prefabrication is not a new concept, but it has come a long way since its inception about 70 years ago, and it’s gaining traction. This fact is illustrated in the article, “10 Basic Facts You Should Know About Modular Homes.” Single-family homes are just one example of the types of construction that lend themselves to prefabrication; others include education buildings, hospitals, apartment buildings, and student housing.

Buildings are more likely to be completed on time and within budget when prefabricated in a factory setting, and worker safety increases due to less debris, less construction traffic, fewer trades working simultaneously, and less activity conducted in proximity to the general public.

For these reasons and more, maybe prefabrication is in your firm’s future. The leading prefabrication tools in the industry are from eVolve MEP. Request a demo of eVolve MEP today, and prepare to embrace the changes. 

Leave a Reply

Your email address will not be published. Required fields are marked *