How Machine Learning is Improving Construction

A look at how BAM Ireland uses Construction IQ to reduce project risk.

The construction industry is massive. People all around the world need buildings to live in, work in and relax in. As more people join the population, more buildings will be needed. With 8.6 billion people estimated to inhabit this planet by 2030, we’ll need to build an average of 13,000 buildings every single day to accommodate everybody. That’s an order as tall as a skyscraper—or a few thousand of them.

Last week, Autodesk held its annual Connect and Construct Summit in London to discuss how the industry can tackle the growing demand. One of Autodesk’s many construction partners at the event was the Royal BAM Group, a lifecycle construction firm with projects around the globe. Mere miles from the Summit at Glaziers Hall, BAM is working on the redevelopment of King’s Cross, a 67-acre industrial site that’s being turned into a community of homes, shops, schools and office buildings.

Part of the King’s Cross redevelopment in London, England. (Image courtesy of Autodesk.)

Part of the King’s Cross redevelopment in London, England. (Image courtesy of Autodesk.)

One of BAM’s most interesting recent innovations is the use of machine learning to enhance its workflow, which it highlighted at the Autodesk Summit and subsequent Autodesk University London.

Too Much Construction Data

With a turnover of around USD$9.1 billion (€8 billion), operations in over 30 countries around the globe, and 50 or more active projects at any given time, BAM collects a lot of construction data. Too much data, in fact. The company had collected so much data about its projects that it even hired a data scientist to sift through it all.

The company hoped it could use that data to its advantage, catching errors sooner and fixing problems quicker. It isn’t just a matter of convenience. In the construction industry, the later a problem is discovered, the more it costs to fix. If a problem is found in the design phase of a building that costs $1 to fix, that same problem discovered in the construction phase will cost $20. If the building reaches the operational phase before the problem is discovered, it’ll cost $60. Clearly, finding problems earlier in the process can save construction companies like BAM a lot of money.

The cost of fixing problems in the construction business rises exponentially as a project progresses. (Image courtesy of BAM Ireland.)

The cost of fixing problems in the construction business rises exponentially as a project progresses. (Image courtesy of BAM Ireland.)

BAM began to better understand its data when it started using Construction IQ, part of Autodesk’s BIM 360 platform. Construction IQ uses BIM project data and machine learning to predict risks and identify issues of safety, scheduling and cost. For the first time, BAM’s wealth of data could be put to use.

“[Autodesk] really opened our eyes,” commented Michael Murphy, BAM Ireland digital construction operations manager.

Leveraging Construction Data with Construction IQ

Construction IQ taps into BIM 360 data to identify the highest risks associated with a project. It bases its risk assessment on safety issues including fall hazards, water hazards, pending inspections or whether or not an issue is overdue. Construction IQ then categorizes subcontractors based on the number of high risk issues associated with them.

In this way, BAM soon discovered a surprising fact. It wasn’t any of its subcontractors that had the highest associated risk—it was BAM itself.

“We realized that we were the biggest source of risk to our projects,” Murphy said.

One of the reasons BAM was judged to be so high risk by Construction IQ was simply due to the number of issues that workers had failed to properly close. Workers would fix problems but leave them open in the system because the value of the digital bookkeeping wasn’t apparent. Properly accounting for what’s fixed and what’s a real risk allows Construction IQ’s machine learning algorithms to provide actionable and realistic insight, and this insight soon helped BAM remove itself from the list of high risk subcontractors.

Once it had its own house in order, BAM could look at the high-risk subcontractors it managed and help them to improve their own safety. For BAM, it’s not a question of judging those subcontractors identified as high-risk (those who build glass houses…), but using data insights to help all its partners improve.

Screenshot of Construction IQ. The bottom left panel shows a list of subcontractors and their associated risk level. (Image courtesy of Autodesk.)

Screenshot of Construction IQ. The bottom left panel shows a list of subcontractors and their associated risk level. (Image courtesy of Autodesk.)

Construction IQ helped direct BAM in prioritizing its attention by pulling focus to the most urgent issues. With an interface that BAM praised as simple and intuitive, the analytics software reveals which issues take priority, which subcontractors need more attention and which projects carry the most risk. The software also identifies water risk issues and safety risks to help avoid costly and dangerous incidents.

“We have an unbounded opportunity to find and address risk,” Murphy said.

Overall, BAM Ireland credits the use of Construction IQ with a 20 percent improvement in quality and safety, 25 percent more time spent focused on tasks and risk items, and workflows that are now 95 percent digital.

Making the Most of BIM 360

Murphy and his colleague Simon Tritschler, BAM Ireland technical deployment specialist, also reflected on how Construction IQ increased the company’s use of existing BIM 360 tools. As Construction IQ began to reveal actionable insight, employees of BAM Ireland began to see more value in using the digital platform. Within six months of adopting Construction IQ, usage of other BIM 360 tools skyrocketed throughout the company.

For example, BIM 360 Field, a construction project management tool, saw 20 percent more active users and 8 percent more active projects at BAM Ireland. BIM 360 Glue, a BIM collaboration tool, saw 24 percent more active users and 13 percent more active projects. And most impressive of all was BIM 360 Docs, a construction document management tool, which saw a 470 percent spike in active users and 103 percent more active projects.

Machine learning, automation, and digitization are becoming ever more prominent. In the construction industry, which lags behind in adoption of these technologies, it’ll be the front runners who define a new era of building.

Written by

Michael Alba

Michael is a senior editor at engineering.com. He covers computer hardware, design software, electronics, and more. Michael holds a degree in Engineering Physics from the University of Alberta.