Towards Cloud Informed Robotics

Sven Stumm, Peter Neu and Sigrid Brell-Cokcan
Pages 59-64 (2017 Proceedings of the 34rd ISARC, Taipei, Taiwan)

The current construction industry slowly adapts towards the new possibilities created through digitalization. However there still a major gap in the information flow between planning phases and the start of construction. With modern robotics we see a chance to mediate this gap. By using cloud accessible CAD software and linking it directly with a web interface, new means of cloud informed fabrication become possible. Designers can easily obtain feedback regarding recurring issues or design demands. We want to enable users to create their own design based on rules and parameters preselected by the designer. Therefore processes for individual parts require an integrated design. Considering changing user demands and the resulting usage requirements the optimization in the phase of conversion from the user´s perspective can help the design process. A number of fabrication technology services exist, which allow for mass customized product creation. The required web interfaces used for these kind of services however are created on a case by case basis. Additionally the required production parameters have to be transferred to the production process. An easy flow from parametric design model towards a web interface for user feedback and fabrication as a service was developed as a first prototype. High quality data is key to automation. Only if the geometry, position and structure of objects are known, automated processing can be planned and executed. The approach we propose within this paper uses cloud bases BIM as the continuous framework for linking design processes directly to production and constitutes a novel strategy. By using cloud accessible CAD & BIM software and linking it directly with robotic fabrication, new means of cloud informed robotics and production is possible. Future developments would allow other robots executing similar task to learn from any other cloud connected robot. Programmers and designers could then also obtain feedback regarding recurring issues during their programs execution. As well as the influences these constitute for the design. A geometric representation and material information of a building element (e.g. a brick wall) must be translated into an individualized assembly process with adaptive machine control strategies. Within this paper, we present a concept for robotic construction processes informed by data collected within a cloud environment. The geometric data, material structure and user feedback etc., stored in the model will be used for automated process and path planning within the digital built environment. As a first proof of concept we show the feasibility of the approach and consider information redundancies as well as missing process information.

Keywords: Cloud, Robotics, BIM, Production immanent design