Recently, I had the pleasure of attending the annual ACADIA conference from 19-24 October. Hosted by the University of Cincinnati, the conference theme was ‘Computational Ecologies: Design in the Anthropocene‘. The conference comprised of two main parts: a 3-day workshop, and a 3-day conference.
There were seven workshops offered in total:
- Open Geo data + Performance (Elcin Ertugrul, Mostapha Sadeghipour Roudsari & Chris Mackey);
- Bio-Agency (Igor Pantic & Soomeen Hahm);
- Dynamo: Intro/Analysis/Optimisation (Colin McCrone & Matt Jezyk);
- Computational BIM in Practice (Konrad Sobon & Nate Holland);
- Prototyping Experiential Future (Shane Burger & Ana Garcia Puyol);
- Kangaroo2: Form-finding and constraint-based modelling (Daniel Piker); and
- Robotic Woodcraft (Sigrid Brell-Cokcan, Johannes Braumann, Daniel Goldbach & Elisa Lublasser).
Open Geo data + Performance workshop
The Open Geo data + Performance workshop was run by the creators of Ladybug and Honeybee. The workshop proved to be fantastic. Unlike some other conferences, there was no pressure to exhibit the outcomes of the workshop, which meant that our time could be dedicated entirely to learning new skills.
Chris Mackey opened the workshop by discussing the notion of tool verse toolkit. Essentially, Ladybug and Honeybee have been designed to be intentionally difficult to use. While this may seem counter-intuitive, the rationale behind this decision is that often with these types of software, users don’t fully understand what is happening ‘under the hood’. This means that often incorrect input parameters are entered, resulting in meaningless results. By allowing access to individual tools (components), users are forced to understand how they work before a result is returned. Furthermore, it allows users to customise their script, hence the notion of tool vs toolkit. It is the programming equivalent of a breathalyser before you can drive.
The workshop then went through several tutorials and computational challenges.
Geographical information systems (GIS)
Using the Grasshopper plug-in ‘@it’, developed by Elcin Ertugrul from Thornton Tomasetti, we learnt how to import open source GIS data for the creation of context massing. While this proved to be interesting, its usefulness is very much tethered to the information available. For well-documented cities where there is an abundance of information, such as New York or San Francisco, GIS works perfectly. However, upon searching for open-source information for Australia, particularly Sydney, we discovered there was little information available. However, since the conference, the Australian government has made geocoded national address data available via data.gov.au. The Grasshopper dataset can be downloaded from here.
Utilising Ladybug’s view analysis components we tested view and overshadowing constrains for Central Park in New Park. The view analysis can be run using either view type or points. A view type is a pre-generated view analysis using either; Horizontal Radial, Horizontal 60 degree cone of vision, Spherical, or Skyview. Download the original dataset here.
As an introduction to Honeybee, we covered the minimum steps required to set up a daylight model for a room. This methodology used the ‘mass2zones’ component to prepare the geometries. Download the original dataset from here.
Building upon the daylighting analysis, we then used the ‘Honeybee_Glare Analysis’ component to run a glare analysis. Note, this process needs the ‘ImageFromPath’ component from the Embryo plug-in. Download the dataset from here.
The basis of this task was to find the most comfortable outdoor places for coffee shops in the west village of New York. This was done by translating the solar radiation values that exist in *epw climate files into mean radiant temperature values. In this way, it is possible to account for solar radiation directly falling on people in comfort. Download the original dataset here.
These are used to filter weather data from the *epw file to identify certain conditions. For example, to achieve cross-ventilation, the optimum dry bulb temperature should be between 18 and 22 degrees, with a relative humidity of less than 80%, and a wind speed greater than 2m/sec. This can be defined as 18<a<22 and b<80 and c >2; and connected to the ‘condStatement’ input of the ‘Ladybyg_3D Chart’ component. This will filter out the entire weather data to only include those analysis periods which comply with the condition.
Heat island effect
Using the newly released Dragonfly plug-in, we learnt how to take building geometry from an urban area and warp rural/airport weather data to reflect the climate of the urban area and its urban heat island. This is done using a heat balance algorithm in an engine called the Urban Weather Generator (UWG). Download the original dataset here.
Ladybug & Honeybee
By far the most valuable thing to come out of the workshop was the ability to engage directly with the creators of these plug-ins. Both Mostapha and Chris are incredibly helpful and willing to respond to user feedback and requests. This can be seen in their generation of the Ladybug Primer and Honeybee Primer to assist with training staff, and secondly, in the development of Ladybug for Dynamo, which is now available on the Dynamo Package Manager.
The conference which ran over 3-days focused on design in the anthropocene, that is, the proposed epoch that begins when human activities started to have a significant global impact on Earth’s ecosystems. Overall the conference had a strong material science and robotic focus. In general, talks were 15mins in duration, with the keynotes lasting an hour. Highlights included:
Pioneers of Design Computation keynote
By Chuck Eastman, Don Greenberg, Tom Maver and chaired by Robert Aish. This special panel was a truly inspirational talk by some of the grandfathers of computational design. Don Greenburg presented some of his research in the field of computer graphics and rendering that he has been working on since 1966. While Tom Maver presented some of his early ideas for computing and computational design. The panel concluded with a call to arms to further the pursuit of computational design. The entire panel received a standing ovation and was easily the highlight of the conference.
Temporal and Spatial Combinatorics in Games for Design
By Jose Sanchez (USC). Jose presented his work on Block’hood, which explores the relationship between architecture and gaming. The game allows players to create neighbourhoods that need to take account of environmental, as well as social and economic, factors in order to flourish. Block’hood gives gamers a range of 96 interdependent blocks from which to construct their cities. These include architectural elements such as apartments and shops, natural features including trees and bushes, and power sources such as wind turbines and photovoltaic panels. Each block has its own range of input requirements that need to be satisfied in order to prevent decay, which can lead to structures becoming abandoned. In this case, they have to be destroyed. Jose won the inaugural Bentley ACADIA Emerging Researcher Award. The full paper can be downloaded from here.
Form Follows Flow: A Material-driven Computational Workflow For Digital Fabrication of Large-Scale Hierarchically Structured Objects
By Laia Mogas Soldevila, Jorge Duro Royo and Neri Oxman (MIT). This was a fascinating talk which presented additively manufactured structures printed from biodegradable and biocompatible natural waste materials. The full paper can be downloaded from here.
Exact Face-offsetting for polygonal meshes
By Elissa Ross and Daniel Hambleton (Mesh). Elissa is a mathematician who has developed a general algorithm for face-offsetting for polygonal meshes that places no restriction on the original geometry. The algorithm uses graph duality to describe the range of possible combinatorial outcomes at each vertex of the mesh. This approach allows the designer to specify independent offset distances for each face plane. The full paper can be downloaded from here.
Innovative Academic Program Award of Excellence
Achim Menges (Stuttgart University) presented some of his work from the Institute for Computational Design (ICD). This mainly consisted of the research pavilions that have been built by students using robotics. Achim was awarded the ‘Innovative Academic Program Award of Excellence’, which recognises ICD’s outstanding contributions to the field of computational design research and education.
Overall the conference was a who’s who of the Food4Rhino website, with the following plug-ins represented:
- Bumblebee (David Mans)
- Dragonfly (Chris Mackey)
- Elefront (Ramond van der Heijden)
- Firefly (Andy Payne)
- Flux (Google X)
- Honeybee (Mostapha Sadeghipour Roudsari & Chris Mackey)
- Kangaroo (Daniel Piker)
- Kuka PRC (Robots in Architecture)
- Ladybug (Mostapha Sadeghipour Roudsari & Chris Mackey)
- Lunchbox (Nate Miller)
- Mantis shrimp (Konrad Sobon)
- Rhynamo (Nate Miller)
- @it (Elcin Ertugrul)
The conference and workshops were a fantastic success, and we look forward to the next one, which will be hosted by the University of Michigan from 27-29 October 2016. To see the wrap-up from the ACADIA 2017 conference, check out this post.