Yuxi Liu is a research associate at the Digital Design Unit (DDU). In 2020, he started to research the use of interlocking system in architecture and worked with the Intelligent Autonomous Systems Group (IAS) at technical university Darmstadt to develop an interface between Grasshopper and Python. This interface can algorithmically aggregate blocks into different configurations by using reinforcement learning. The paper “Reinforcement Learning for Sequential Assembly of SL-Blocks” was adopted by 39th eCAADe Conference.
In 2021, he completed his master degree at technical university Darmstadt with a thesis entitled “SL-blocks for design”. During the thesis, he explores and designs a large variety of interlocking aggregations made from a singular discrete building block, where all the connections of blocks are dry-jointed. The work further explores one coherent construction of SL blocks with different materials assembled into a branching column carrying a roof. His current research focuses on the autonomous robotic assembly using Artificial Intelligence as well as the development of interlocking systems for discrete design and assembly.