The SECOAS Project—Development of a Self-Organising, Wireless Sensor Network for Environmental Monitoring. Available online: http://www.lancs.ac.uk/staff/marshai2/ secoas/SANPA-2004-Britton.pdf (accessed on 3

In this paper, we examine the SECOAS project which aims to design and field a self-organising, wireless sensor network for environmental monitoring. We examine the background of the project and follow the design through to an implementation in field trials. We show how implementation issues led to several significant changes in the direction of the project. We explore one aspect of the project—the operating system kOS—in detail, and explain the rationale for various design decisions. The lessons learnt during this process will be of interest to other researchers and developers in the areas of wireless sensor networks. 1. Background Currently, oceanographic studies frequently involve using large, expensive devices to log data, typically for several months at a time. During each deployment, there are substantial risks that the platform will be damaged or destroyed [2]. This approach also has the disadvantage that the sensors measure environmental phenomena at only one physical location. An alternative is to use a network of sensors to build a spatio-temporal picture of the environment. This leads to a system that is robust even when nodes are destroyed or the network topology changes. Furthermore, nodes can be added at will or reconfigured for various purposes. The availability of increasingly low-cost microprocessors and radio devices has made this approach feasible from an economic point of view. 2