2 days workshop | 06-07th Jul | 15 seats available

This workshop aims at reviewing the learning opportunities in the parametric design process arising from integration of data specific to challenges that need to be resolved on a design project, their analysis, and the automation of the exploration process to maximize learning and generation of knowledge about the design.

We will be using Bentley's GenerativeComponents (GC) with the optimization services framework.  We will consider and include proxy analysis methods into the design process and employ optimization using Bentley's industry-strength genetic algorithm. We will discuss and implement additional analysis types that might deepen understanding of design alternatives; how any parametric design model exploration promotes learning about possible approaches to design solutions and deepens the design team's knowledge about the design; how, therefore, any parametric design model is comparable to a sketch -and needs to be readily discarded once it has been exhausted; how multi-objective optimization increases learning opportunities beyond any single-objective optimization process; and how all these pieces can contribute to better design solutions when used in the design process in smart ways.

We will explore further challenges and opportunities posited by the fusion of more massive data into the design process like building information modeling (BIM) and what “parametric BIM” or “” might be. 

Parametric design tools like GC allow design teams to develop parametric models that are able to generate a huge number of design solutions.  Many of these potential designs are never actually explored because there just is not enough time. Integration of analysis into the design process facilitates deeper exploration of any examined solutions and thus helps designers better understand the potential performance of their designs. These insights mixed with experience and intuition can help design teams accelerate their learning, build up faster their knowledge about the design, while moving it towards better performance. 

This type of evaluation can be automated with the help of optimization algorithms. For multi-objective optimization, genetic algorithms with their mix of goal oriented or fitness-driven and arbitrary processes like mutation and cross-over promise to explore the design space with a greater chance of finding high-performing solutions. Sets of these high-performing solutions often illustrate well what the specific trade-offs are between conflicting design goals. This allows confirmation, adjustment, or growth of a team’s knowledge about how to achieve the design goals, may provide new insights for modification or replacement of the parametric design model so that it embodies the design goals better than the previous version, or even leads to revision of design goals. Overall it increases the depth of understanding of the design and thus increases the chances to improve the design further than the conventional design process.  


Day 1:

  > parametric design principles
  > introduction to working with GC
  > integration of parametric design and analysis
  > multi-objective optimization
  > exploring case examples
  > develop new cases from examples
  > end of day discussion and planning day 2

Day 2:

  > parametric BIM– – demo/exploration and discussion
  > case examples continues – alternative performance analysis types
  > discussion of cases
  > case examples continues – collecting design artefacts for presentation
  > final discussion

Software required

  > GenerativeComponents (GC)
  > GC Analysis Add-in
  > GC Optimization Add-in
  > MS Excel [optional; but helpful].

All Bentley software used in this workshop will be provided to participants as downloads prior to the workshop.

Hardware required

Participants need to bring their own computers capable of running Windows OS 7 or 8 and GC with the add-ins.  This is taxing software; therefore, mobile workstation-grade laptops are recommended.  It is recommended that the latest Windows Updates have been applied to the computers and preferably all software has been installed before the workshop starts.