b'IN THIS INDUSTRY, WHEN WE LOOK AT THE COMPLEX WORK THATMECHANICAL CONTRACTORS HAVE TO DO, THE HUMAN ELEMENTWILL STILL HAVE A BIG ROLE TO PLAY. KEN LANCASTLEto collect, assess and synthesize projectOne of the most compelling applicationsconsiderations going into their solutions and business data on a much greaterof AI in construction is for enhancing theare way more sophisticated than anything scale. At the same time, AI platforms andhealth and safety of the human workforce.AI is able to do now.algorithms are being adopted to automateOrganizations are already investing inIn fact, he continues, there is a case recurring manual tasks (aka machineknowledge-based systems that enableto be made that AI will make human learning), thereby freeing human workerscompanies to use data from past incidentstalent more valuable. Ajay Agrawal is the to focus on more engaging and mission- to better predict project risks, embedGeoffrey Taber Chair in Entrepreneurship critical tasks. proactive measures and create moreand Innovation and professor of strategic What AI is doing now, at least from oureffective health and safety programs andmanagement at the University of Torontos perspective, is helping to address some ofstrategies. Elsewhere, companies are alsoRotman School of Management. In the more programmable and repetitiveexploring AI-driven systems that can scanhis book, Prediction Machines, Agrawal tasksthe ones that can be brokenjobsite video feeds and images to detectuses economic theory to suggest that by down into a specific and predictablesafety risks as they emerge and triggermaking one part of the process easier and sequence of events, says Ken Lancastle,preventative measures. more cost effectivee.g., through AIchief operating officer for MCAC.From optimizing the estimating andother people involved in the process will By allowing AI to play a role, there isbudgeting processes to predictingrise in value, rather than decline.potential for contractors to free up staff toschedules and mitigating project risk,So, rather than thinking of AI as do other things that maybe require morethere are many back-office uses for AI.something that is competing 100 per cent innovation or a human touch. Moreover, down the road, we can expectwith workers, Agrawal makes a point that, Consider the estimating process, heto see AI assume a more physical presenceno, its going to take a thing you do and continues: As an example, one area whereon the jobsite with robotic equipment.make it so cheap that everything else you there is potential for AI would be theThis equipment can be programmed todo is going to be more valuable because review of specifications and the provisiontackle repeatable, time-consuming or eventheres more aggregate demand,of a baseline estimate from that data.riskier jobsite tasks, once again freeingsays Seaton.Mind you, that estimate is only as goodhuman workers to tackle things that make as the data its made from, but you canbetter use of their skills. In short: the robots are not coming to see the potential. From that point there isThe Human Workforce Isntreplace us. At least, not for a while, and more opportunity for review and Going Anywhere not in a way that will negate the demand human eyes. Despite the considerable benefits of AI,for skilled human talent.On the topic of data, AI offers contractorsthere is a reluctance among tradespeopleWere always going to need the human the ability to locate and connect key datato fully embrace the AI revolution. Afterelement, says Lancastle. Are there going from disparate data sources that mayall, many ask, if automated systemsto be some jobs that AI can take away? I otherwise be forgotten or lost within and autonomous equipment can do thethink the safe answer is yes. But in this an organization. job, where does that leave the humanindustry, when we look at the complex Theres this idea that 95 per cent ofworkforce? Its an understandable concern,work that mechanical contractors have to construction data is locked up in someand yet, the truth for now is that the AIdo, the human element will still have a big PDF somewhere. This is called darktools and technologies finding their wayrole to play.data, and what AI can do is go throughinto the trades today are designed more Youd be hard-pressed to find an all that dark data and extract informationto help human talent rather than industry that isnt talking about artificial from it, explains Seaton. For example,replace them. intelligence. And as more self-learning contractors can go into their dark dataTheres a concern that if AI gets too goodtechnologies and systems find their way and say, I want to be able to look at everyat what it does, then people lose theirinto contractors offices and jobsites, RFI and run analyses against them. Andjobs, but we dont see that happening anythe conversation is becoming louder instead of that being a long and manualtime soon, says Seaton. Contractorsin construction. Time will tell how AI process, they can use AI to get into thosearent just executing tasks from a book;ultimately impacts each trade, but for data sources and make the process theyre creatively solving problems allnow, there are compelling reasons to much easier. the time, and the skill, experience andwelcome the AI revolution.NATIONAL BULLETIN | Issue 2, 202321'