Machine Streams V1.0

Welcome to our first Machine Learning webinar series, Machine Streams V1.0. This 4 part, approx 16 hour, Rhino + Grasshopper webinar series will be taught by DesignMorphine’s Zvonko Vugreshek. Having the machine take on more decisions than the user has been ever so popular, it’s implemented through almost all segments of life nowadays. Whether you are designing for an art show or producing shop drawings for your local workshop, you are dealing with some form of optimization of choice along a couple of parameters. The optimal choice is often difficult because we are limited as beings to make such complex decisions in a rather short amount of time. Even when writing these lines. That is the key point where we introduce machine learning, to help people make better decisions and thus reducing their potential errors. Examples from facade structuring, roof panelization and urban design will be shown and answered using various methods and models ranging from single objective optimization to neural network classifiers. 

The webinar series will take place across 4 days on 2 weekends in July, 2020. Each webinar will be approx. 4 hours long with live Q&A’s.  Webinars will take place in a Private Facebook group via Facebook Live and video uploads. Participants will be able to ask questions by commenting on the videos, this gives the webinars a more interactive community environment. Each webinar will be pre-recorded and uploaded at the time of the webinar to ensure optimal quality and to avoid issues of connectivity that can occur with live streaming. At the live stream times, Zvonko will be live streaming to summarize the videos and answer questions. All videos will remain available, including the live streams, for your viewing as much as you like, even after the live stream. This means you can also enjoy the videos if you are not available at the live stream time. You may view them at your convenience and availability. All webinars will be conducted in English and there is no previous experience of the topics required. You must have a valid non-business Facebook account to attend and watch the videos. 



Dates: Jul 18, 19, 25, 26 (2 weekends)
Time: UTC (Coordinated Universal Time) 2:00 PM (for each day above)
Duration: Approx. 4 Hours each (approx. 16 hours total for all 4 parts)
Location: Facebook Private Group (videos will remain available in the group to view as often as you like)
Tutor: Zvonko Vugreshek (see bio on bottom of page)
Software: Rhino, Grasshopper (Plug-ins: Lunchbox ML, Pufferfish, Mesh+, Heteroptera, Weaverbird, Octopus, Silvereye, Owl, Wallacei)
Requirements: A valid non-business Facebook account 


Part 1 – July 18, 2020

Intro to optimization and ML:

We will start off with a gentle intro lecture as there are quite a few things that are better explained this way rather than jumping straight into rhino/gh. Afterwards we will test out some of the said methods and techniques on basic examples in grasshopper that would set the foundations for the next session.


Part 2 – July 19, 2020

The art of simple choice(s):

Using the methods, we got introduced to last time, we are going to look around on how to establish algorithms that can make efficient simple or binary choices/classifications for us from a couple of inputs, for example any site or massing data.


Part 3 – July 25, 2020

Sort, group, select -> generate:

From the choices from last time now we need to evolve into more complex scenarios. Now handling multi-dimensional inputs and processing them to produce meaningful data for generating a form based on the specific relationships between the given inputs.


Part 4 – July 26, 2020

Feasibility wrap-up:

Having the form in the optimal place means its ready to be turned into a meaningful structure. For this we will synthesize everything learned in the previous sessions adding some construction and feasibility details in our optimizations along the way.



Machine Streams V1.0
Zvonko Vugreshek
Macedonia & Croatia
Member @ DesignMorphine
Research Assistant @ BTU Cottbus-Senftenberg
Zvonko Vugreshek is a Croatian/Macedonian Architect and Computational Designer, currently a Research Assistant @ BTU Cottbus-Senftenberg. He has implemented these techniques since his mid-bachelor studies and is among the pioneers in introducing them in greater depth to his colleagues at the faculty of Architecture at UKIM Skopje. His interests are creating tools for creating optimized and effective designs regardless of the scale using site-specific factors such as environmental, social and behavioral factors. He has worked for Copenhagen studio 3XN where he was responsible of developing computational tools to solve project-specific tools and automate tedious tasks, resulting in a range of in-house components for everybody to use. He has been part of quite a few DesignMorphine workshops, lectures and projects so far.



Facebook Private Group