Leiden University Pushes Generative Design
Leiden University is hosting an upcoming event on "Deep Generative Models for Engineering Design," signaling a growing academic focus on advanced computational design. These machine learning-based tools are being explored to help generate, test, and iterate design options for the built environment more rapidly.
The academic push into generative design aligns with Dutch national goals for a fully circular economy by 2050, with an interim target of a 50% reduction in primary resource use by 2030. The construction sector is pinpointed as a priority area in the government's transition agenda, aiming to make circular construction the norm. This involves not just reusing materials but also designing buildings for future adaptation and deconstruction from the outset. This digital approach supports the decarbonization of the Dutch building sector, a critical task given that over 90% of buildings historically relied on natural gas. The government's goal to transition 1.5 million homes off natural gas requires major renovations and new heating systems. Generative design can optimize for energy performance and material efficiency, contributing to the national target of reducing greenhouse gases by 55% by 2030. Practically, generative design is already being applied in the Netherlands. The Dutch construction company Van Wijnen partnered with Autodesk to create a generative tool for urban planning, optimizing residential layouts based on metrics like solar gain, views, and cost. In Amsterdam, the 70-meter hybrid timber tower "Elements" used a generative algorithm to balance KPIs such as daylight, wind comfort, and energy performance, resulting in a structure with over 50% lower CO₂ emissions than conventional construction. The use of advanced digital tools extends to Digital Twins, which are being implemented by Dutch municipalities to improve policy and planning. The Brainport region, for instance, uses a digital twin to model housing scenarios and their impact on traffic and air quality as it plans for 62,000 new homes by 2040. These initiatives are supported by the national government, which sees digital twins as crucial for integrating data and managing social challenges in the physical environment. The Association of Netherlands Municipalities (VNG) is actively exploring the role of AI in urban planning, recognizing its potential while also developing governance frameworks to ensure responsible application. The VNG is also supporting the development of a Dutch-language model, GTP-NL, to increase the transparency and reliability of AI used by municipalities. This reflects a broader trend of Dutch universities, like TU Delft, embedding digitalization, AI, and data-driven design into their core urbanism research programs. This push is contextualized within broader European Union policies. The European Green Deal and the Renovation Wave initiative emphasize digitalization as a key enabler for achieving climate neutrality and enhancing circularity in the construction sector. EU strategies promote digital product passports and shared data infrastructure to create a more efficient and sustainable building industry across member states.
