While attending the MCA convention this year, I got the opportunity to talk to Bob Snyder of Binsky and Snyder Mechanical Contractors. Bob is about to start working with LINQ, an artificial intelligence program that acts like a digital assistant focused on construction information. I have been working with LINQ for several months in our VDC and service areas, and as a CTO my focus has been information search and availability. As an executive, Bob had a different focus. Bob was intrigued by the idea that LINQ could understand questions. Not just what we mean when we asked the question, but what information is likely to answer it. In Bob’s mind the true value of AI in construction would be in understanding which questions to ask about a project even before information exists to answer those questions, so the AI can inform us when information becomes available to answer questions that effect our critical paths.

As Bob talked about the value of questions, one of the more fascinating ideas that he shared was the idea of doing a project kickoff that produced questions rather than answers. In this version of the kickoff meeting project questions are the focus. Bob’s theory was that if we could define the questions ahead of time for LINQ, it could be looking for answers right off the bat. Bob also hoped that digital assistants down the road would be able to also help us evaluate the most valuable questions and recommend questions to us based on previous installations.

In looking back, I think of this concept as the Binsky kickoff, and even without the help of the digital assistant, I can see it becoming a valuable part of any project. I’m certainly familiar with the critical nature that good questions play in getting the project information we need, but I think after talking to Bob that I underestimated the value of the question itself to a project.

I often hear people talk about not knowing what they don’t know and about the blind spots on a project. Having the people typically involved in a kickoff meeting forced to ask only questions offers a new way to find these blind spots. The Binsky kickoff resembles more of a workshop, the kind used in design thinking and product development. These workshops attempt to identify problems in need of a solution by focusing on the issue rather than then answers. In doing this, they provide a process to separate perceived problems from those that exist but are less obvious — the blind spots.Question Marks Balance Scale Image by Wm. T. Spaeder

But even once we’ve identified all the questions, evaluating them is likely to be a challenge, and this will probably be the place digital assistants help us most. Not all questions are created equal when it comes to construction. Determining what are the truly valuable questions, the ones that if answered have the biggest impact on success, will be hard to do, and likely be a moving target. While there are plenty of books and articles on the proper questions to ask before, during, and after a project, they tend to be very general, and are not the nuts and bolts questions of an actual project but rather useful generalizations. Processes like Six Sigma’s The Five Whys can have significant value but are not the critical questions for a specific construction project. A digital assistant’s understanding of the detailed questions that make up the execution of a project is likely to be immensely more helpful. With the introduction of tools like LINQ capable of collecting those questions, we may finally have the chance to see what is actually being asked on projects, not just in the planning stages but in the final execution. With those questions in hand we have the chance to prioritize information that satisfies the most common and pressing questions first.

Cell Phone Question Image by Wm. T. Spaeder

Beyond prioritizing questions, digital assistants may help us improve the quality of the questions we ask in the first place. I think it’s fair to say that questions can vary dramatically in clarity and detail. Each stakeholder has questions that are critical to success in their aspect of the build. Understanding these questions as a team is key to being able to answer them. Often a request for information is missing key details. For instance, if I ask you for floor drain information, the information I want will depend on whether I am buying the floor drain, drawing the floor drain, installing the flooring, or doing maintenance on it. To get the answer right, you need more information — or for me to ask better questions. In working with LINQ, I found I occasionally had to ask a question several different ways before I got what I was looking for. As I did that, LINQ probably learned how to better at answer my questions, but more importantly I learned how to ask my questions more clearly. That’s important in and of itself. Too often, we blame poor RFI responses on bad information or lack of understanding when really the problem is that the question was unclear. Perhaps unintentionally, digital assistants give us a way to improve the clarity of the questions we ask by offering us instant feedback, so that we can adapt and learn to ask better questions.

Use of processes like the Binsky kickoff to help focus on questions before a project starts, and the help of digital assistants available to track the questions throughout the project lifecycle could provide a new metric to measure successful projects and understand them. In a previous article on Single Source of Truth, I focused on the ability of tools like LINQ to be truth seekers for a project. I talked about how they could look for the answers that best suited the questions we ask. After talking to Bob, I realize that beyond finding answers digital assistants are likely to change the landscape of the questions being asked in the first place, helping us to ask more valuable questions in a clear way. This will offer new insight into what makes up a successful project so we can better answer the big question of how to build better projects.

Jonathan Marsh

Leave a Reply

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