SailAway: Spatial Cognition in Sea Navigation

Pedestrians, bicyclists, car drivers, boat and airplane pilots, as well as other cognitive agents participating in public traffic must respect rules in order to avoid dangerous situations and to ensure a smooth flow of traffic. The SailAway project [1] investigates traffic and related navigational rules from a formal and computational point of view. The aim is to enable artificial cognitive agents to act in compliance with such rules. Traffic rules, which are expressed in natural language, usually subsume distinct, but similar situations and actions under more abstract spatial or temporal concepts and relations. In this paper we describe an approach to representing rules that exploits this qualitative nature of natural language descriptions used in traffic laws. Based on this approach we present methods for determining actions that are rule-compliant with respect to the current spatial situations of the agents involved. Finally we present the prototype of a control system of boats in sea navigation that implements exactly these methods.