At the sold out Design Technology Summit July 17th, 2019 in Seattle Washington, there will be discussion of Artificial intelligence (Ai). Ai is something that will surely feed the accelerating phenomenon of perpetual innovation our industry is experiencing. You will know from previous articles that the AEC Industry is seeing an unprecedented increase in the number of tools available to us and an increasing number of firms developing their own tools through scripting in-house. Before Ai really gets into the mix, there’s an important step that needs to be made in any industry to prepare it. Data collection. As dry as that might be, is critical for any sort of intelligence, even human intelligence, to be effective. In the case of Ai, any data we will want it to rely on will need to be collected and in a format it can “ingest”. Simple concept perhaps but if you look at current sources of data common in firms, the organization is highly varied. The outputs are inconsistent. Tying systems together is challenging at best.
Ai will need to be “spoon fed” data in digestible chunks to be effective. The information it uses is critical. Garbage in, garbage out applies. But for the AEC industry, and for AEC firms, where will this information come from and how is it best composed? The answer to these questions all depends on what’s being asked. For example, if a designer wants to know how annual wind patterns might affect a 200-foot high-rise, then she would want data from NOAA. If a Project Manager wants to know how to most efficiently staff a large multi-phase project then she would want project staffing, productivity, project change order data, bill-rates, and related data collected within her firm from past projects. The results of each of these are reliant on the quality of the data input. The recognition of the potential cost savings mined using Ai may be profound. It may be the proponent of significant change to how buildings are delivered. It’s uncharted territory.
As our industry pioneers into this new territory with platforms unknown, it makes sense to begin thinking about the questions that we will want answered and almost more importantly, the data with which it will be informed. Perhaps that data collection should begin now. How translatable is our data?
If interested in participating in next year’s Design Technology Summit, please send your meesage to [email protected]