Generative design: human expertise meets artificial intelligence

Illustration of the design variants created with generative design

How STRABAG is using algorithms and generative design to optimise building design

Generative design is a key component of STRABAG’s efforts at transforming the construction industry. The use of big data and corresponding algorithms not only makes it possible to generate various deductions and forecasts but also to pursue specific goals in the design. This is precisely the aim of generative design. The model-based design method offers building planners countless possibilities for designing their structures. And generative design unfolds its full potential thanks to the combination of artificial intelligence and human expertise.
Generative design offers STRABAG a design tool for leveraging various data potentials that make construction not only more innovative, but also more sustainable. The possibility to influence resources, quality, costs and deadlines is greatest in the early planning and design phases of a construction project. Generative design, unlike previous practice, offers not just three or four design drafts, but a whole multitude of them. This saves time and resources in the search for the optimal design solution – fast, automated and data-based.

Generative design: a principle in three steps

Generative design works with automatically generated data. In the computer-aided design process, evolutionary algorithms and optimisation methods generate a large number of design alternatives in a very short amount of time. These are created by combining the design variables of a parametric model. The principle of generative design can be explained in three steps:

Graphic of the three steps of generative design

Hundreds of design proposals are generated in a very short amount of time. Innovative results are created that might have remained undiscovered with manual design planning. 
The system analyses the design variants with regard to the optimisation goals. Design specifications, including the defined parameters and spatial conditions, are automatically taken into account. 
The result: optimised design proposals based on the previously defined requirements and goals. The most suitable variant is then chosen and further adapted if necessary. 

Overview of tools and functions

Determines the heat load, cooling load and air volumes – in real time. Stored climate data records can be used to determine the photovoltaic (PV) yield and optimise the PV surfaces. 

Analyses and optimises the individual parameters of a building in terms of carbon emissions and production costs in the design.

Automatic generation of a counter-rotating, two-part precast staircase with support.
Optimises the classic design and planning process in the design of excavation pits.

Generative design in use: decarbonisation of existing buildings with GD ENERGY

The STRABAG planners are already using generative design in the ongoing Z2 construction project at the ZÜBLIN Campus in Stuttgart. The planned refurbishment of the existing building offers numerous opportunities for using digital tools, especially GD ENERGY.

GD ENERGY allows STRABAG to parametrically model the existing building and evaluate its energy efficiency right from the initial design and planning phase. Based on data from Google Earth and using the escape and rescue route plans, the planners determine the design parameters such as air volumes and heat and cooling loads in the preliminary design. The building’s potential PV yield can also be generated using the stored geographical data (GIS) and compared with the primary energy specifications in accordance with the Buildings Energy Act. So by using generative design, STRABAG is helping to create a climate-neutral, future-proof building.

Grafik eines Gebäudes, auf der von GD Energy errechneten Energieträgern abgebildet werden
The photovoltaic module in GD ENERGY calculates the possible energy yields. 

Outlook and next steps

STRABAG currently uses four generative design tools. The next step is to link these tools even more closely with each other in order to make the entire planning and design process more interactive. The goal is to be able to view the building as a whole and across several planning and design disciplines simultaneously.

One of the pilot projects for the application of the GD framework in the context of networked planning is the Innovation Center at the ZÜBLIN Campus in Stuttgart. From the design planning to the structural design to the office space design and building services design, everything here is automatically networked on a data basis.

Abbildung des Innovation Centers in Stuttgart bei dem generative Verfahren zum Einsatz kamen
Video: The Innovation Center in Stuttgart is a pilot project under the GD framework.

This might interest you as well.

  • Tablet showing surface simulation.

    Use of augmented reality in road construction

  • Use of augmented reality in road construction.

    Digitalisation, Processes and Innovation

    More digital, more innovative, more efficient.
    Explore our projects