Action Selection in Everyday Activities : An Opportunistic Planning Model Based on Bounded Rationality, Minimization of Effort, and the Effective Use of Space

One of the main problems of performing ill-defined tasks is how to decide what to do next, i.e., the problem of action selection. This especially applies if tasks are not or only weakly constrained, such that the each action could be done in any order, equating a multitude of possible solutions. In order to gain a deeper understanding of the cognitive processes involved in human action selection, which in turn can inform the development of artificial cognitive agents such as robots, this thesis focuses on everyday activities as a subset of ill-defined tasks.

Previous research shows that even if no hard constraints exist, people exhibit specific preferences for action orderings, which arise from bounded rationality, i.e., having limited knowledge and computational power (e.g., in the form of working memory) available. As a result, people aim to minimize the required overall physical and cognitive effort. In the context of spatial tasks, this can be achieved by taking properties of the spatial environment into account and using them to one's advantage and by employing a stepwise-optimal action selection strategy. This thesis presents the Opportunistic Planning Model (OPM), an explanatory cognitive model that instantiates these assumptions.