Abstract
The insufficient level of reproducibility of pub-lished experimental results has been identified as acore issuein the field of robotics in recent years. Why is that? First ofall, robotics focuses on the abstract concept of computationand thecreationof technological artifacts, i.e., software thatimplements these concepts. Hence, before actually reproducingan experiment, the subject of investigation must beartificiallycreated, which is non-trivial given the inherent complexity 5.Second, robotics experiments usually include expensive andoften customized hardware setups (robots), that are difficultto operate for non-experts. Finally, there is no agreed uponset of methods in order to setup, execute, or (re-)conduct anexperiment.To this end, we introduce an interdisciplinary and geograph-ically distributed collaboration project that aims at implement-ing good experimental methodology in interdisciplinary roboticsresearch with respect to: a) reproducibility of required technicalartifacts, b) explicit and comprehensible experiment design, c)repeatable/reproducible experiment execution, and d) repro-ducible evaluation of obtained experiment data. The ultimategoal of this collaboration is to reproduce thesameexperiment intwo different laboratories using the same systematic approachwhich is presented in this work.
Links and resources
Tags
community