Drawing Fabrication


The workshop presents a cyber-physical workflow, where users can instruct a robot through on-object-hand-drawn notations [1,2]. An integral part of the workflow is a machine learning pipeline and decision tree that enables the robot to see, interpret, and carry out given actions autonomously based on a physical drawing. This breaks all current modes of digital fabrication, which commonly uses a digital model as the starting point, from which numeric information is extracted. This workflow aims to alter the use-paradigm of digital fabrication, enabling such systems to be used for sketching or thinking besides production.

The workshop explores the inner workings of the workflow, by allowing participants to develop their own drawn semantic robot language and decision tree. Participants will use bespoke tablet interfaces to develop a drawn semantic language, the decision tree is developed in Rhinocerous3D(Rhino) and Grasshopper3D(Grasshopper), where a custom toolbar allows participants to access Machine Learning functionalities from the ML,net library. The developed drawn semantic language will be tested using a Universal Robot (UR) equipped with a pen.

The workshop can allow for a maximum of 15 participants, that will be divided into groups and given a tablet with a dedicated “Drawing Fabrication” app which allows to save the drawn semantic language.

The workshop grants participants first-hand experience with robots, while learning about Machine Learning, Rhino Grasshopper workflows and are allowed to question current methods of numeric robot interfacing.

Technical requirements:
All participants is expected to bring their own laptop and a working Rhinocerous3d license. Basic to intermediate Rhino/Grasshopper skills is a plus, but no robotic fabrication experience is required.

[1] Pedersen, Jens, Narendrakrishnan Neythalath, Jay Hesslink, Asbjørn Søndergaard, and Dagmar Reinhardt. ‘Augmented Drawn Construction Symbols: A Method for Ad Hoc Robotic Fabrication’. International Journal of Architectural Computing 18, no. 3 (1 September 2020): 254–69. https://doi.org/10.1177/1478077120943163.
[2] Pedersen, Jens, Asbjørn Søndergaard, and Dagmar Reinhardt. ‘Hand-Drawn Digital Fabrication: Calibrating a Visual Communication Method for Robotic on-Site Fabrication’. Construction Robotics 5, no. 2 (1 June 2021): 159–73. https://doi.org/10.1007/s41693-020-00049-2.

Workshop Leaders

Jens Pedersen, Odico Construction Robotics, Aarhus

Jens is an architect and inventor. He is currently carrying out his industrial PhD, which researches workflows related to onsite robotics and timber construction. As part of his studies, he has developed a novel construction system and a patent pending interface method for robotic fabrication. Prior to his PhD, Jens has worked in practice internationally, while organizing workshops as part of the AA_Visiting School program or in the form of student workshops at the Aarhus School of Architecture.

Kristoffer Gade, Odico Construction Robotics, Aarhus

Kristoffer is an architect and skilled computational designer with extensive experience in modelling and fabricating complex geometries through parametric workflows. Kristoffer has taught and helped organise student workshops during his studies at the School of Architecture Aarhus.