Thursday, August 7, 2008

PS 66-133: SensorKit: An integrated sensor data collection system for field ecologists

Fabio Silva1, Eric A. Graham2, Annette DeSchon1, Yeung Lam2, Thanos Stathopoulos2, Wei Ye1, Jeffrey Goldman2, Terry Benzel1, William Kaiser2, and John Wroclawski1. (1) University of Southern California, (2) University of California, Los Angeles

Background/Question/Methods In recent years, Wireless Sensor Networks (WSNs) have emerged as a promising technological advance in environmental monitoring systems. Even though the potential for WSNs to contribute significantly to ecological advances remains high, their adoption has been overshadowed by fairly steep technical knowledge requirements. To alleviate this problem and promote adoption of WSNs to the ecological community, we have developed a system called SensorKit. SensorKit provides an end-to-end system from the sensors to the database. It includes commercially available hardware (sensors and node platforms), and open-source software, for remote configuration and control, data transport, and storage. SensorKit also includes installation and operation tutorials, which together provide a user with the basic know-how required to deploy a sophisticated embedded sensing system. The SensorKit development model is open and community-based. Our goal is to catalyze a community of users and developers to exchange ideas and experience, identify and provide new capabilities, and provide on-going bottom-up support for current and future configurations and features. Results/Conclusions SensorKit features include adaptive sampling capability; wired and wireless communication between nodes, the controlling computer, and the Internet; compatibility with existing sensing hardware (e.g. dataloggers) as well as analog, digital, and advanced sensors (e.g. imagers); an expandable architecture to support multiple nodes and sensors; solar and battery power operation; deployment optimization tools; and data integrity functionality. SensorKit offers simple instructions, a web browser interface for accessing data stored in a database, and an intuitive custom-designed graphical user interface that allows ecologists to configure data sampling rates and methods (such as average, min-max, or raw data), as well detailed metadata for each channel. In addition to accessing and controlling SensorKit nodes remotely, an ecologist in the field is able to make configuration changes, perform sensor calibration, and check data integrity by simply using his/her laptop or a PDA device. SensorKit fills a niche for individuals or small teams of researchers and practitioners who have demanding sensing requirements but do not have ready access to engineers and computer scientists or large cyberinfrastructure investments at their study sites. These investigators require powerful, robust sensing systems for systematic studies in their domains and a simple path to deployment and analysis so they can focus their expertise on discovery rather than the enabling technology.