Automatic real-world assembly of machine-designed structures
Luzius Brodbeck, Fumiya Iida
- 发表年份
- 2014
- 引用次数
- 7
摘要
Several approaches have been presented which allow robots to build structures to adapt themselves or their environments. To autonomously build these structures, a design must be made, from which instructions for the fabrication process can be derived. For a constrained fabrication process, e.g. considering the limited range of a robot, this transfer can be cumbersome. We present a local building process based on a sequence of two distinct operations, which implicitly encodes the shape of a structure. Given this encoding, the structure can readily be built with a real-world robotic system. We show automatic design of structures reaching out of the robot's range and fulfilling stability and strength constraints using an evolutionary design algorithm. The final design can then be built with a robotic arm from wooden cubes and hot melt adhesives. We demonstrate the whole process including the construction of a structure from more than thirty cubes with our real-world setup. We expect that automatic design and construction can further improve the physical adaptability of robotic systems.
关键词
相关论文
Statistical Learning Theory
Yuhai Wu, Vladimir Vapnik
1999
Artificial intelligence: a modern approach
1995
Fractional Differential Equations
Igor Podlubný
2025
Applied Nonlinear Control
Jean-Jacques Slotine, Weiping Li
1991