The Pitfalls of Skipping Submittal Register Reviews
June 6, 2019
When someone thinks about using Artificial Intelligence (AI), they probably picture a robot with all of the capabilities and common sense of a human, but with the precision and speed of a high-end computer. It’s tempting, given this interpretation, to assume that the AI automatically does the work you assign it without any input or effort from you. This, however, is a risky assumption to make. Yes, AI can help you perform a task faster, with greater accuracy, and much more ease, but it still requires some human touch. This is why we always recommend that you thoroughly review any submittal log, product/equipment list, and/or closeout requirement data, etc., regardless of how it was generated.
The reason for this lies in the data used to generate those lists mentioned above. In some cases the design team will often out-source the writing of project specifications to another party. These third parties tend to recycle specifications and product data from previous projects, copying and pasting old specs and turning them into “new ones.” As a result, by the time the project team gets their hands on the specifications, they can be bloated and full of possibly irrelevant information. This can include things like products and equipment that are no longer manufactured, extraneous meetings with the design team or subcontractors, standards that are out of date, and manufacturers who worked “their magic” in order for their products to be listed at the exclusion of other options.
This all goes back to the old adage of garbage data in, garbage data out. When you upload specifications full of garbage to an AI platform, it unfortunately doesn’t apply your common sense and experience working on construction projects to parse out the junk (not yet at least, we are working on it). This is why reviewing a submittal log generated via AI and machine learning is critical to ensuring the success of your project.
Now that we are in agreement over carrying out a review, how can this be efficiently accomplished? No one really wants to read hundreds to thousands of pages of dull construction specifications. You will be tempted to utilize ineffective shortcuts like searching for keywords or skimming spec sections, rather than reading the whole thing. If someone actually does attempt to read the specs in their entirety, I’ll bet they make it 50 pages before their eyes completely gloss over and the exercise proves to be pointless. But, seeing that you are smart and used an AI-based submittal generation platform, you can take advantage of the time savings by looking for impactful and high-priority information contained within the specifications. This can include, but is not limited to, reviewing and answering the following questions:
- Do the specifications align with the contract documents and drawings? Is anything potentially out of scope?
- Are any divisions or spec sections missing submittals or product data?
- How many shop drawings are called for? Does this make sense for the size and scope of the project?
- Has the architect delegated any design elements to the subcontractors?
- Will the project be undergoing LEED certification? If so, what is required?
- Do the requested mock-ups make sense for this project?
- Are specified materials and products going to need to be imported? If so, what are they and where can you acquire them? How long will they take to deliver? Are there tariffs active on those materials?
- Do any products/equipment have a sole source designation? What are the specified products with a basis-of-design requirement? Which materials will have a long-lead time or require custom fabrication? Can we start preparing for this before mobilizing?
- Are there QA/QC requirements that call for very specific inspections and certifications?
Carrying out a proper review greatly increases the chances of finishing your project on time, on budget, and with as little litigation as possible. With profit margins currently sitting around 3.5% to 5%, a few things missed here and there can really affect the P&L for a project. Factor in the possible cost of claims and litigation, doing a little analysis of the specifications will go a long way to make sure you’re not on the wrong end of a losing project. Or worse, a lawsuit.
AI and automation platforms like AutoSpecs should act as a safety net in your workflow. You can rest easy knowing that the bulk of a task has been taken care of, but with so much at stake, and the time freed up to do it, why not have someone review your log after it’s created?
Want to learn more about submittals?
Watch our webinar for submittal education and best practices.