Putting AI to Work Mitigating Project Risk
July 30, 2020
The risk of missing a requirement can weigh heavily on the minds of project teams. Slim profit margins leave no room for error, and there are countless tasks that project teams are responsible for juggling throughout the lifecycle of a project. Missed requirements often mean added risk, stressful project disputes, and even the potential for litigation down the road. With such high stakes in mind, a typical project team is still resource constrained when it comes to actually building a project. Which means that there are few man hours actually available to devote to reading and re-reading large construction documents to comb for anything that might have been overlooked.
We sat down and interviewed Pype’s Director of Client Success Josh Matheny, and Document Crunch’s Co-Founder and Construction Industry Curator Josh Levy, to discuss how artificial intelligence is leading the way for solving this problem.
How do construction companies make money?
Josh Matheny, Pype: The general contractor tries to estimate how much the project will cost then adds a small profit on top of that. After they complete the project, if their estimate was right, they take home their expected profit. If their estimate was wrong, any budget overruns eat into their profit and they end up taking home less money than they expected, or can even end up losing money on a project. This is only looking at the pursuit side, not factoring in any building delays, weather issues, injuries, change orders, or the multitude of other potential issues that might arise.
Josh Levy, Document Crunch : There’s a lot of financial risk involved for the contractor, obviously. This is a very risky industry, which typically yields single-digit percentages of profit. One or two mistakes can quickly take a project into the red. Because of that slim margin of error, one of the highest priorities for contractors is mitigating risk. And much of the risk exposure to the contractor lies in the contracts that it has entered into for a project, which includes terms and conditions, specs, drawings, etc.
How do contractors mitigate risk?
Matheny: The easiest thing a contractor can do to mitigate project risk is to have a full understanding of the information contained in the project documents. This includes knowing their rights and responsibilities as written in the contract, knowing what their insurance policies cover (and don’t cover), and knowing everything that is required in the specs and plans to complete the job. It is far easier and far better to avoid mistakes in the first place than it is to fix them.
Levy: And that’s why many blue-chip companies have invested heavily in their legal departments, in insurance compliance functions, in precon and estimating departments, and other governing bodies that look to assess and mitigate project risk.
If companies are investing in personnel to mitigate risk, why use AI?
Levy: Smaller and mid-size companies typically don’t have the volume of work or profitability to justify making such a large overhead investment at their current size. However, their smaller volume also means that any litigation or financial disruptions they encounter will likely have a greater impact on their bottom line, far more than a large volume company encountering one single dispute. The smaller companies just don’t have the cash flow to absorb an impact of that size, and that’s one of the biggest reasons general contractors go under.
Matheny: Even at larger companies, these documents are so massive and so numerous that you could dedicate an entire team to reading them and extracting important information, but there are so many other tasks that need to be done that require a human touch. What artificial intelligence and machine learning platforms can do is take high-risk, monotonous tasks and automate them, then present that information to humans to interpret and use effectively on the project.
What do Pype and Document Crunch do?
Matheny: We both analyze construction documents, but our difference in scope is in which documents we analyze. Pype’s AutoSpecs and SmartPlans solutions look for submittal requirements, closeout requirements, product data, schedules, and other critical project requirements in spec books and drawings, respectively. We can then export this information as a submittal log, as draft product, equipment, and finish schedules, closeout logs, and procurement logs. We also recently added a new algorithm called Pype AI to both platforms. It compares your uploaded specs or plans to millions of similar sets and industry best practices, then flags what requirements are likely missing from your project documents so you can discuss them with your design team before it impacts your budget or timeline.
Levy: Project teams, and even those supporting project teams like seasoned executives, have little to no patience for reading through legalese and boilerplates in contracts and insurance policies—they’re there to build. However, an understanding of such “rules of engagement” is crucial to leveraging successful outcomes at the project level. Document Crunch analyzes those documents for them, and organizes the information into an index we call “the Crunch.” Here, project team members can explore the documents by provision type, letting them quickly see which types of provisions are or aren’t in your documents, easily jump to those provisions in the document, and understand what those provisions are for in layman’s terms. This can be used for some of the high risk legal terms, but also for every day management terms like, how to give notice, what constitutes a delay event, how substantial completion is defined, etc. We have also started creating automated workflows where we auto-populate contract checklists using such algorithms.
As AI platforms like Pype and Document Crunch continue to emerge in construction, contractors should be reviewing them critically in comparison to their current solutions and workflows, knowing that even one mistake can tank a project’s projected profits. Project teams interested in these software solutions for risk mitigation can visit https://pype.io and https://www.documentcrunch.com to learn more about their specific platforms.