Research

This project deals with the problem of teaching robots how to execute a wide range of meal preparation tasks. For this, we first focus on the task of Cutting Food. We propose to access Web knowledge to teach robots how to perform a variety of cutting actions on a range ob objects. On this website you can get more information and an idea on how a robot would parameterise action plans to execute meal preparation tasks.

In the research lab you have access to and can experiment with generalized robot plans for transporting objects within man-made human living and working environments, including apartments and retail stores. The generalized robot plan controls different robots transporting a variety of objects in different environments for different purposes in different contexts.

This research project addresses the challenge of instructing a robot agent to learn novel tasks interactively, specifically in the household domain. The focus of this repository lies on the task of teaching the pouring task by using written instructions with PyCRAM.

For more information, you can visit the webpage of Interactive Task Learning to get a better idea on how a robot can learn from different teaching methodologies.

TIAGo robot in the IAI Bremen apartment laboratory.

Joint Probability Trees (short JPTs) are a formalism for learning of and reasoning about joint probability distributions, which is tractable for practical applications. JPTs support both symbolic and subsymbolic variables in a single hybrid model, and they do not rely on prior knowledge about variable dependencies or families of distributions.