Knowledge Graph Laboratory Interactive Example
In this virtual research lab, we aim to empower robots with the ability to transform abstract knowledge from the web into actionable tasks, particularly in everyday manipulations like cutting, pouring or whisking. By extracting information from diverse internet sources — ranging from biology textbooks and Wikipedia entries to cookbooks and instructional websites —, and linking it to scene graphs of the environment as well as motion parameters of robot action plans, we create knowledge graphs that inform generalized action plans. These plans enable robots to adapt their body motions in accordance to cutting techniques such as slicing, quartering, and peeling, as well as inferring tasks for various fruit properties and using suitable tools . Through this, they make abstract web knowledge practically applicable in robot perception-action loops.