Optimizing Design Decisions With Machine Learning

This workshop focuses on using machine learning to make design decisions. It covers theory and practical applications in various areas, such as facade and urban.

Duration:
8 Hours
Difficulty:
Beginner
Language:
English
Certificate:
Yes
Registration:
€105.00
Members:
89.25 EUR (28.60%) discount
Recordings:
Available Indefinitely

Optimizing Design Decisions with Machine Learning

Nowadays, computers are often the ones calling the shots, and this trend is everywhere. From creative/aesthetic decisions to preparing blueprints, you’re always trying to make the best choice based on a few key details. But making the right call quickly isn’t easy because there’s much to consider. That’s where machine learning and artificial intelligence come in—tools that help us make smarter choices and reduce potential errors. Now, they are accessible more than ever, especially for designers, engineers, and planners.


The Scope of the Workshop:


The course will primarily focus on the technical aspects of using machine learning for design decision-making. During the course, we will cover both the theory and practical applications of machine learning to optimize parameters in different design scenarios, such as facade, roof structuring, and urban design. We will explore a range of methods and models, including single-objective optimization and neural network classifiers, to demonstrate how machine learning can be used. While the primary focus will not always be on the aesthetic outcome, we will work on optimizing the most valuable parameters required for the current example.

  • How to recognize optimizable problems in your workflow or tasks and automate them.
  • First steps into data science and machine learning
  • Analyze existing morphologies/structures and optimize them according to specific criteria.
  • Make design decisions based on computed recommendations.
  • The workshop will adopt a hands-on approach, combining lectures and practical exercises. Students will work individually on design projects that optimize various design parameters using machine learning and AI methods.


  • The workshop will cover the following workflows:


  • (Design) problem identification and formulation
  • Data collection and analysis
  • Model selection and utilization.
  • Optimization and evaluation
  • Design decision-making based on computed recommendations.

Program:

Day 1:


  • Lecture about ML/Optimization/Data Science for design decision making.
  • Hands-on design of a structure to optimize.
  • Project assignments.
  • Q&A

Day 2:


  • More advanced concepts and building up on the projects/topics from the previous day.
  • Project development and reviews.
  • Q&A




  • Grasshopper 3D
  • Rhinoceros 3D
  • LunchBox
  • Owl
  • Octopus
  • Total sessions: 2 Sessions
  • PAACADEMY will provide a certificate of attendance.

Course Content

Instructors

0 Followers

0 Following

Biography
Zvonko Vugreshek is a Berlin-based computational designer, digital fabrication specialist, and data scientist passionate about pushing creative boundaries and will deliver a 90-minute-long exciting tutorial at the upcoming Computational Design: NEXT 14 online conference. He has been involved as a research assistant at the Digital Design & Fabrication department of the Brandenburg University of Technology in Cottbus as well as a teacher at the chair of biodigital architecture and sensorics (CyPhyLab) at the Technical University of Berlin. He has worked at offices for all scales of design and engineering, most notably at 3XN, doing optimization and fabrication of large projects, such as the Sydney Fish Market. Zvonko has been applying Machine learning and automation techniques in all scopes and scales of design and engineering. He co-owns the Berlin-based generative design and fabrication studio Pixolid, bringing his experience in AI/ML to the physical world. He has been involved in many workshops, webinars, and lectures aiming to popularize computational practices.

0 Comments

You must be logged in to comment.