Self-Adaption in Lego-Mindstorm Train Control System using OSGi

Abstract

Public transport has become an important part of everyday life in today's society. With the advances in information and communication technology has ITS become an important research field. Because of this, we have in recent years seen an increasing number of autonomous transportation systems. These autonomous systems operate in dynamic environments where unpredictable events may occur. These events may affect the vehicle's ability to operate safely. It is therefore essential that these systems have a way to reason and react to these events. Since the vehicles often operate under high speed is it important that adaptation to these events can happen quickly and while the system is running. This paper presents an adaptation module design intended for autonomous trains that operates on a Lego Mindstorm based model. The module's task is collect contextual information and use this information to adapt the trains behavior accordingly. To enable this adaptation during runtime is the Java based framework OSGi used. The module uses a state machine that is implemented with the State Design Pattern. This state machine is used to as the context reasoning component for the adapter. OSGi based services are used to facilitate retrieval of the contextual information and to perform the adaptation actions. The module is implemented using the modular development tool Reactive Block and was tested on the Lego Mindstorm based train model mentioned above with good results.