RESEARCH OVERVIEW

Through this research project, a novel deposition method with bound granular materials was explored. The goal was to develop a behavioral robotic process that can fabricate customized components out of aggregate materials. An integrated process consisting of a material system, a multi-task end-effector for 6-axis robots, and a communication protocol was successfully developed.

‍

DEVELOPMENT CONTEXT

This investigation challenges the state-of-the-art systems where specific aggregate behaviors can be achieved through unbound tailored particles. The production of these elements usually requires the use of resource intensive fabrication and specialized formwork (eg. in the case of plastic). In contrast, the use of a readily available product with a natural material was proposed.

‍

MATERIAL SYSTEM

Wooden toothpicks were chosen as the test material for aggregation due to their high availability and low cost. UV Resin was used to bond the loosely aggregated granules in a manner of seconds. Bound granules allow a higher typological variation than simply aggregate granules. Initial experiments were conducted manually to gain insights into the material and its manipulation.

‍

FABRICATON SETUP

A KR 125 KUKA robot with a multitasking end effector grabs toothpicks from a box, places them on a fabrication surface, and binds them with UV resin. The unpredictable nature of toothpick aggregation provides an opportunity for a non-determined fabrication process, as the next drop-off point to continue building a structure needs to be decided on the go.

‍

END-EFFECTOR

The end-effector incorporates all the necessary functions for fabrication. This includes a gripping system, a spraying system, and a UV Light for resin curing. The gripping system consists of three “claws” activated by individual servo motors. In the spraying system, the correct processing of resin became critical as the material could be cured accidentally with UV light and sunlight. This system consisted of a pressurized container with a built-in compressor, a pressure sensor to adjust the spraying pressure, and two solenoid valves for activating the spraying and venting of pipes. A nozzle cover moved by a servo is activated before turning on the UV light to prevent the clogging of the system.

‍

COMMUNICATION PROTOCOL

An Arduino board controlling the end-effector, a webcam for sensing, and a laptop as a control center were used for physical communication. The laptop was connected to the webcam and Arduino via USB and to the robot via ethernet. PyCharm IDE, Grasshopper, and the KUKA system were used for establishing communication between the physical devices. Through PyCharm the webcam’s images were processed and converted into contour information with the OpenCV library. This was sent as point data to Grasshopper via socket with the gHawl plug-in. In Grasshopper, the Simulacrum plug-in was used for the intercommunication with the KUKA robot and with the Arduino board via Firefly for controlling the end-effector.