I’m Vignesh Kaushik. I curate and write articles on Thank God It’s Computational to help architects, designers, and urban planners leverage cutting-edge technologies on AEC projects.
TGIC is a community-driven platform providing concise technology education to students and professionals in the AEC industry to learn at their own pace. If you’re not a member, consider joining the community and supporting our work.
Machine Learning for Architecture & Urban Design
How many meeting rooms do you need in an office? It’s a simple question, but one that is very difficult to answer. WeWork employed machine learning to assist in forecasting how often meeting rooms would be used. The most powerful implication of this study is that before they begin construction, the design team can plan a space that fits the needs of the members that will one day occupy it.
Neural network outperforms a human in predicting meeting room usage
Google have piloted use of ML models at multiple data center facilities and have produced a 40% reduction in energy used for cooling and 15% reduction in overall energy overhead. If you want to learn more on how Google is using neural networks to optimize data center operations and drive energy use to new lows, here is a white paper (PDF).
Take a bunch of data, find the hidden interactions, then provide recommendations
Many healthcare projects on the boards today will not be finished until the early 2020s. Are we designing buildings that support the way our clients will deliver care on opening day? Three teams of architects and planners were asked to envision how artificial intelligence (AI) and other advances in technology will shape a healthcare setting in 2025.
Start-ups use sensors and machine learning to do “predictive maintenance”, spotting faults in building systems like heating and air-con before they crash. Data from sensors built into HVAC units paired with a machine learning algorithm predicted 76 out of 124 real faults, including 41 out of 44 where an appliance’s temperature rose above tolerable levels, with a false positive rate of 5%.
A unique combination of crowdsourcing and machine learning was used to build a system that can provide users with parking difficulty information for their destination, and even help them decide what mode of travel to take. In a pre-launch experiment, there was a significant increase in clicks on the transit travel mode button, indicating that users with additional knowledge of parking difficulty were more likely to consider public transit rather than driving.
Since science, technology and creativity have a long, intertwined history, selecting which metaphors to explore is an important research decision. This article explores three metaphors: Augmented Creativity, Computational Creativity and Creative Systems.
There is a growing awareness about machine learning and the many ways it can shape our immediate future. Unfortunately, several misconceptions have also grown up around it, and dispelling them is the first step to understand ML better.