Severe energy limitations, and a paucity of computation, pose a set of difficult design challenges for sensor networks. Recent progress in two seemingly disparate research areas namely distributed robotics and low power embedded systems has led to the creation of mobile sensor networks. We conjecture that augmenting static sensor networks with mobile nodes addresses many design constraints that exist in static sensor networks. The possibility of combining sensing, computation, communication and actuation to not only passively monitor the environment (like static sensor networks) but also actively track, and in some cases mitigate problems, makes it an interesting area of investigation.
We present here the Robomote, a robot platform that constitutes a single node in a mobile sensor network. The design goals in building the robomote has been twofold: - Maintain compatibility with the Berkeley mote, a popular platform for research in sensor networks. It is our belief that maintaining compatibility with the mote helps easy initiation of use of the Robomote. - Minimal capabilities in order to minimize the cost of the platform to enable large-scale deployment and testing.
We describe the software and hardware design of the second generation Robomote. We then list two case studies where the new robomote has been used to study algorithms afor mobile sensor network. Finally, we conclude by listing some of the challenges in enabling large networks of such mobile sensors in reality. These include rethinking some of the decisions taken in static sensor networks and some that are new to mobile sensor networks.