CLAM: a Time-Sensitive Approach to Mobile Robot Localisation And Mapping

A precondition for a mobile robot to autonomously explore its environment is the ability to self-localise. When a map of the environment is not available, the robot should be able to self-localise and, at the same time, to build a representation of the environment. Here the problem: position estimation needs an environment model, and world modelling needs the robot position. Current research proposes solutions based on the simultaneous execution of the two activities. We propose a novel approach based on their concurrent execution. This approach, titled CLAM (Concurrent Localisation And Mapping), is founded on the conjecture that a proper separation of concerns may help in breaking the loop of the problem. Localisation and Modelling, acting on different time scales, are mostly independent each other. When a synchronisation is needed, it is controlled by an external and suitable strategy. Finally we are testing our approach by exploiting the Real-Time Performers framework based on a new set of architectural abstractions modelling the temporal behaviour of a system.