In a recent post, I mentioned that I’ve been asked to help lead our efforts in a particular area of research. I was deliberately vague at the time. Now, with Autodesk University 2022 around the corner, it seems a good time to say more about this. I hope it stimulates some interesting discussions at the event.
As a company, one of Autodesk’s core interests is to enable the design and creation of healthier, safer, more resilient spaces. We want to explore tools for designers and architects that help them create buildings with more attention to the occupants. In the past, architects achieved this through their own intuition and adherence to building codes and standards that support the occupants’ needs. In recent years, the AEC industry is pushing to involve occupants earlier and more often in the design process. This human-centric approach to building design will still consider cost, schedule, energy and more, but creating healthier spaces for people will take center stage.
Our team of researchers intends to encourage human-centric design by providing better tools. This could include providing better indicators for human-centricity like comfort or safety. We will define “soft metrics” – relating to things like comfort, well-being, habitability and perhaps even productivity – that might be integrated into a traditional design process. Focusing on human-centric building design will also include exploring tools that encourage and streamline broader stakeholder and occupant feedback during the design process. Perhaps this means using existing and future generative tools that support more interactive and collaborative workflows.
There’s a lot of supporting technology and exploration that will go into this effort. For simulation, we’re looking to harness spatial analysis toolkits such as VASA to assess proximity and visibility. Also, we expect to use ontologies to define and access semantic information about buildings. With sensing, we’re looking at capturing data from our Living Lab in Toronto to validate assumptions and to learn more about occupant behaviours. We’ll probably use digital twin technology such as Project Dasher to visualize the data. We want to use reinforcement learning along with other ML techniques to explore emergent occupant behaviours by placing intelligent agents in a variety of spaces. We also want to support the creation of spaces that evolve over time, whether via near-term re-configuration or as the needs of successive generations of occupants shift.
In case you’re interested in learning more about this effort at Autodesk University 2022 in New Orleans, please drop me a note and we’ll find some time to chat. (We can also connect separately if you’re not attending, but at this stage that may not happen until sometime in mid-October.)