Enabling Smart Navigation by Incorporating Human Inspired Evaluation of Route Sections

Abstract

It is challenging for every child not becoming lost in the shopping center. Simulated agents have the same problem. So they need some localization. On the basis of the existing cognitive architecture in th project Simulation of Mental Apparatus and Applications (SiMA) at the Institute of Computer Technology (ICT) currently, we show, how cognitive architectures can get extended by location awareness and navigation in a human-like way. This paper deals with particular requirements for the used data structures to enable self-localization and pathfinding in a bionically inspired way. The typical data structure for navigation in the form of a connected graph gets modified and extended to be compatible with psychoanalytic concepts and still allow navigation. The functionality was implemented exemplary in a multiagent simulation.

Functional Integration of Human Like Localization in a Decision Unit for Autonomous Agents

“…It has turned out that developing human-inspired navigation and self-localization is just the halfway of having a simulated agent behave human-like. We continued to the investigation, how unconscious thinking influences the human navigation in [6], which lead us the work at hand. So, we have all the developed methods aligned with the existing model, ready for implementation.…”

“…After further discussion in the project team, it became clear that the primary process has for sure influence on the navigation process. This influence was investigated in [6] in detail. Now, as there is a stable view of how human-like navigation should work, we have aligned the described navigation-and selflocalization-methods to the functional modules of the SiMA model.…”