Arbitration based framework for design of holistic multimodal human-machine interaction

Abstract

Creating dynamic cognitive systems, such as human-machine systems with the appropriate interaction inside, the system engineer is usually facing a high overall complexity. For instance, designing a modern driver assistance system, the developer is dealing with the diversity of drivers’ behavior, the complexity of the traffic and vehicle dynamics and the heterogeneity of already designed assistance systems. To deal with this complexity effectively, the creation of holistic, stable and well-usable cognitive systems implies the usage of a design framework, which could provide benefit and orderliness to the creation process. In this contribution, we propose such a design framework that consists of the theoretically derived generic cognitive system architecture as well as of the arbitration-based methodology and tool-based interface for the interaction design. Besides the theoretical background, we describe the general rules of the framework in a stepwise manner using an automotive example of a newly developed driver assistance system for cooperative lane changes in a highway scenario.