At Henderson Engineers, it’s in our nature to seek the most innovative and efficient solutions for our clients and to constantly try to improve our process. One design concept that’s been a natural shift to propel our projects to the next level is generative design.
Generative design is essentially an iterative design process that uses a computer system or program to rapidly explore solutions using set project constraints, making it possible to explore a variety of design alternatives faster and more efficiently than if a designer had done them by hand. It pulls in existing data from Autodesk Revit and then uses that data as a “foundation” for computational design and optimization. Using these calculations, generative design can then produce multiple design options, select one option from a sort list, and populate it with real Revit elements to help save time and money.
Historically, generative design has been used on architecture and construction projects, but engineering firms have recently started to apply it in design. Henderson is one of the few firms actively using this technology for building systems design to unlock possibilities that were previously unrealistic. As Henderson’s Director of VDC/BIM, I get to spend my days helping solve client problems using virtual design and construction. Here’s an example of how we applied generative design to solve an HVAC design challenge for a client on a tight timeline.
THE CHALLENGE: Balance Capacity, Cost, and Aesthetics
Recently, a confidential client tasked us with minimizing the size and quantity of ducted, forced-air HVAC systems in a cooling-dominated climate – while maintaining a clean aesthetic – with pricing and scheduling constraints. While our primary goal was to meet the client’s needs, we also had to coordinate with other trades to ensure our solution was compatible with their proposed layouts.
One approach our team considered was using radiant cooling panels – specifically Price chilled sails – but they had to align with the structural bay aesthetic per project requirements. Some bays were square, and some were odd shaped because of the building’s footprint. The design team considered three panel sizes to fit the geometry of the bays: 2×4, 1×8, and 4×4. You might be thinking, “How hard is it to optimize a square space with rectangles?” Normally this wouldn’t be a major issue, but in this example there wasn’t a linear relationship between the size of the panel and the cooling load. We also had to consider price and aesthetics, making it more than an engineering problem. Plus, the lighting and fire sprinkler systems had to go through the radiant panel layout. We had a collaborative design problem to solve, and generative design provided a solution.
THE SOLUTION: Options, Options, Options
In generative design, the heavy lifting shifts from the designer to the design programs. The speed in which the layouts are produced, in conjunction with the supporting design data, makes it possible to expedite solutions and pass on cost savings to clients. In this example, the team used Autodesk Generative Design, Dynamo, and Revit to balance the panel type, placement, and rotation – and we got hundreds of possible configurations.
From there, we filtered and sorted the options by coverage area, number of units, cost, and cooling capacity, which allowed us to narrow down the top solutions for the client and architect to make an informed decision. We were able to present the max load offset solution to meet the initial design concept, as well as optimized alternate options such as lowest cost, and provide layouts for a clean aesthetic. Each bay didn’t require the same layout, which allowed the architect to use the cooling panels as part of their design. The speed and accuracy of the algorithm also allowed us to quickly coordinate with other trades.
THE FUTURE: Endless Possibilities Through Innovation
We’re entering an era where common automation and parametric design won’t take our projects where we need them to go. Generative design builds on our already strong design foundation by incorporating computational design data and then optimizing those configurations to solve the project challenge.
As this technology continues to evolve, I will be sharing regular updates around how Henderson is using generative design, including updates on how we’re incorporating it into machine learning algorithms and simulation-ready digital twins. We’re also energized to see the possibilities for energy modeling optimization and integrating BIM into a generative design approach.
Check back soon for more insights.
Join our email list to get the latest design innovations, technical content, new projects, and research from Henderson’s experts delivered straight to your inbox.