The driving concept behind Paint Var Proxy was utilising the disparate technologies and systems we have been developing over the course of term two focused on the application of sensors, robotics and real time feedback for creative visualisation. This was achieved by extracting data that was processed using state of the art methods to the project with the assistance of an industrial robotic arm.
The inspiration for this creative approach to hardware was artists and collaborators that have previously utilised the characteristics of industrial robotic arms for works other than traditional automated manufacturing. The act of using the defining features of a robotic arm, speed, accuracy, strength and patience, allowed us to generate many variations of the many variables included in the setup of the project.
Key technologies used to adapt and create an output were sensing with a Kinect 2, point cloud generation and segmentation. Tool paths for the robotic arm were automatically generated within Grasshopper using Python scripting to streamline the process. Digital outputs and an arduino microcontroller were used to control and synchronise the triggering mechanism to expel paint. KUKAVARPROXY was installed on the KUKA KRC2 controller in order to send the generated tool paths, removing any requirement for manual development.
Moving forward, we foresee continuous development incorporating live data from sensors of different types to allow robotics to play a bigger role both within factories, but also on building sites. The continued research in maximising the development of online programming using robotic arms without hard-coding the tool paths opens up a large variety of possibilities not seen within their original intended use. By experimenting with a creative output, we were able to identify an exciting and new perspective on robotics that will more than likely grow into different approaches towards the technological skill set that is robotics.
youtube link:
Students: Riccardo Mura, Ardeshir Talaei, Hazal Yilmaz
Faculty: Angel Muñoz