self-learning drawing machine made from scrap wood

In my diploma thesis, I dealt with how the design of machines will fundamentally change with the help of sensor technology and algorithms.

In simple machines, the expectations of the machine (such as precision) are mainly represented by the mechanics - for example, by the corresponding motors and bearings. But with the help of simple sensing and neural networks, we have the opportunity to radically rethink this.


Based on the previous work in the workshops, I developed a drawing machine that can be rudimentarily assembled from scrap wood and motor units I developed.


The kinetic structure starts to move while a camera computes the current tool position together with the motor positions and slowly understands how the cryptic structure works.

In the end the machine was capable of creating rather precise drawings regarding the size and lack of precision in the mechanical setup.

You can find more in-depth information about the motor setup here. Also, prior to this project I developed the workshop "the reorchestration of things", that teaches students how to interact electronics and let them build drawing machines – the basis for this project.

Copyright © 2024 Maximilian Hans. All Rights Reserved