BUILDING AND NAVIGATION STRATEGIES FOR ON-SITE ROBOTIC CONSTRUCTION
Building construction typically relies on a centralized design, in which information is used to produce the most appropriate sequence of assembly actions which are assumed to serve as the unambiguous blueprint for building execution. However, uncertainties in actions which lead to violations of tolerances, true dimensions of fabricated modules as well as changes in the environment can require frequent replanning, the cost of which may be prohibitive for construction projects which have reached a level of complexity that becomes increasingly difficult to plan. For similar reasons on-site assembly is often prawn to purely reactive and manual operating sequences.
The central goal of this research project is to master the increasing level of complexity for on-site construction of high-performance building systems by decentralized planning strategies which compute smart assembly sequences based on AI reasoning and thus enable new processes in the research area of on-site large-scale robotic construction. This research project contributes significantly to the research network and demonstrator as it is the only project within the cluster with a focus on high-level planning for building execution.
Prof. Dr. rer. nat. Marc Toussaint
Machine Learning and Robotics Lab (IPVS-MLR), University of Stuttgart
Prof. Dr.-Ing. Dr. h.c. Oliver Sawodny
Institute for System Dynamics (ISYS), University of Stuttgart
- Hartmann, V. N., Oguz, O. S., & Toussaint, M. (2020). Planning Planning: The Path Planner as a Finite State Machine. Workshop on Planning and Robotics (PlanRob).
- Hartmann, V. N., Oguz, O. S., Driess, D., Toussaint, M., & Menges, A. (2020). Robust Task and Motion Planning for Long-Horizon Architectural Construction Planning. Proc. of the IEEE Int. Conf. on Intelligent Robots and Systems (IROS).