Description
We present an experiment in which intuitive stability measurements are tested in 3D space. The motivation for testing this is as follows: In everyday life it is crucial to have some understanding of physical properties and processes for even seemingly simple tasks like cutting food, setting the table or even placing a cup onto a saucer. For this somewhat intuitive physics it might be beneficial to look at how humans behave. Additionally, when human and robot work together, it is crucial for the
robot to adopt these intuitive human heuristics. Otherwise, humans may become uncomfortable and feel the need to
correct the robot’s actions, hindering the development of a productive co-working environment. Even though something is stable considering a theoretical definition, it might not appear as such to the human observer.
Different examples of setting the table in a stable but not necessarily ideal way
Human intuitive physics is an active area of research. In this field, the paradigmatic task of stacking cubes - or assessing their stability - has been utilized in various forms and levels of complexity. This includes virtual reality (VR), towers with a lot of cubes or even different shapes. In spite of this, the most basic setting of this canonical task with only two cubes has not been thoroughly investigated. Additionally the majority of reports feature only a reduced, one dimensional decision space with respect to the stability. In the assessment of stability participants have been found to have a bias towards safety - stacks of cubes which are barely stable in a static environment are deemed unstable. The prevailing hypothesis for this bias is a security margin, separating the area deemed safe or stable from the area yielding collapsing towers. Since reports are based on the reduced decision space it is interesting to explore how this explanation transfers to more realistic settings.
Author and Contact Details
- Nikolai Bahr
Tel: +49-421-218-64294
Email: nibahr@uni-bremen.de
Profile: Nikolai Bahr