I was working on a target proposal for a high performance building just last week, and was reminded again of the surprisingly long list of performance metrics that building designers are faced with today.

Energy efficiency has been in the conversation for many years now. Over time our industry has come up with a lot more specific performance metrics for different aspects of a high performance building.

The list above states the most common performance metrics that I have come across. These are only a dozen of the metrics compared to the many different decision that get made during the process of a building design. How do we know that we are making the right choice?  Enter. Multi-objective Optimization.

Multi-objective optimization is a great way to analyze and visualize the different trade-offs of the various performance metrics. This optimization is used to provide feedback to owner, designer and engineers about the different strategies used to satisfy certain performance targets.

When we run a multi-objective optimization, a lot of times not all the generated alternatives make sense from a design stand point. With this, we can use interactive diagrams to narrow down our search. Think of it the same way we filter out our search for a flight on Expedia. 

Here, I can filer out by # of stops, airline and departure time. In a similar way, once we have some constraints on the design, like, orientation, budget, glazing %, we can filter out options that satisfy these filter. Some times, we are unaware of exactly what these filters are for a particular design problem, so we can use interactive visualizations that help us filter out options on the go.

One such visualization is the parallel coordinate plot below. Where we can adjust the different sliders to narrow down the the design problem. This helps us look at a few thousand options at a time, and narrow down to the  alternatives that meet our baseline filters and performance targets.

Living in the world of big data and multiple options, the architecture industry needs to move out of a single target optimization approach to a more holistic multi objective optimization.

About : Pattern r+d, founded in 2014, enables the design of sustainable and healthy building environments and communities by developing and applying innovative tools, processes and education. Pattern delivers highly integrated design solutions that go beyond shallow notions of sustainability to true parametric and performance driven design.