Identifying the Best AEC Software to Exceed Business Goals
August 25, 2020
Content originally published in Forbes Technology Council.
An investment in AI or machine-learning software can require a lot of time, money, and effort for a business. As a result, your evaluation process of any given tech can be the difference between a positive ROI and flushing limited resources down the drain. The key to any evaluation process of AI, however, is understanding what AI is and how it relates to your needs.
Just as physical tools and machines automate physical tasks, AI automates mental tasks. Both of them supplement an existing workforce, but still require human operators and human intelligence to be of proper use. Expecting to phase out your human employees for a server room full of AI software is as much a recipe for disaster as a GC replacing their subcontractors with a box of nail guns and drills. AI programs are a tool that will make a human workforce more efficient, not obsolete.
How the software improves the efficiency of your workforce depends upon two factors, both of which you need to measure: Client Purpose, and Developer Purpose. The first metric, Client Purpose, concerns your business needs. What problem are you trying to solve? Where are the inefficiencies in your process? Generally, these can be boiled down into three main categories: Speed, Accuracy, and Ease.
As a general rule of thumb, most workflows require you to pick two of those three attributes. Accurate and Ease will mean that your team is slowly and methodically going through the process, taking their time with it, and this naturally will require a lot of lead time. Speed and Ease will mean that it gets done lighting fast, but only because your team was taking shortcuts, skipping steps, and potentially producing half-baked results. Speed and Accurate means long days, working through breaks, and a high likelihood of burnout as soon as the process is over. You probably already have a preference for one of these combinations, and have based your workflows off of them. However, the right AI software can provide that lucrative third piece of the puzzle.
If you’re missing speed, then of course a computer will be faster than a human if only through the way it processes information. Think of it like clockwork—a computer has all of the gears connected to each other, so that when you turn the first one, every other gear turns simultaneously. The human mind, however, has to go step-by-step, gear-by-gear, turning each one manually. Accuracy, then, would be determining the arrangement of the gears and the relevancy of each. An AI program has a set selection of gears, and manipulating the program lets the user rearrange them, but after initial set up they will always remain the same. Humans, however, have to re-determine each gear in the sequence every single time, reinventing the wheel and leaving the process wide open to variation (ie, error). Ease, simply, is the difference between doing all of that work yourself, or having the AI do it for you.
Once you’ve figured out your purpose for finding an AI solution, the next step is actually finding that solution. Every platform will tell you that they can solve your problem and that they’re the best choice—not doing so is counter-productive to being a business. Because of that, it can be difficult to determine which solutions are actually effective, and which ones aren’t. This is where the second metric, developer purpose, comes in.
However, this time you’re gauging the thought-process of the developers who created this software. Generally, there are two main paths that developers can use when creating an AI tool. The first is exploratory, with the developers stumbling upon a neat trick they can do with some computer code, and then trying to find problems they can throw that computer code at. The second is the problem-driven, where a developer encounters a problem and custom-builds a solution for it. While the exploratory process has the potential to adequately address your needs, it’s a lot more of a gamble than problem-driven solutions. Often, exploratory developers will attempt to retrofit their program to solve a myriad of different problems, becoming a jack of all trades but a master of none. And, even if they did find a perfect-fit for their program, these developers are in it for a good time, not for a long time, resulting in little-to-no integration potential, and bug fixes and updates being few and far between (if they happen at all).
A problem-driven solution, on the other hand, will be focused on solving the issue it was made to address. These developers are looking to build best-in-class software and relationships with other businesses. The investment of time and energy into these platforms by the developers will return dividends in the form of custom integrations, rapid bug fixes, and user-feedback-based feature updates. An enterprise partnership with these companies will only magnify all of these benefits, as the platform improves year after year while your monetary investment stays the same, getting more bang for your buck.
At the end of the day, “getting more bang for your buck” is what any investment is about. For AI software, companies like Pype with a proven track record of positive ROIs, regular product updates, and partnerships with key players in the AEC industry will always be a more sound investment for your business than someone looking to make fast cash. AI has the potential to become a critical piece of your project lifecycle, but the wrong tool for the job will only slow everyone down.