Networked Sensor Platform for Largescale Data Collection

J.A. Bert, E.D. Cocker, C.L. Chua, P. Faultley, S. M. Nonis, E.A. Ronchetto, M. Shreve, Z. Landvater, C. Paulson
PARC, California, United States

Keywords: IOT, sensor platform, edge computing, distributed sensing, satellite communication

We are developing a low cost, robust sensor platform capable of recording and reporting data and surviving in remote harsh environments for up to one year. The platform is entirely self-contained, powered by solar panels and backup batteries, with a large array of sensors, a compute module for onboard data processing, and a satellite modem for data reporting. The platform architecture is flexible and can be customized for specific missions. The complete data record is stored locally and processed by the compute module, which runs anomaly detection algorithms to identify activity in the area. Anomalous data, as well as baseline environmental data, are reported to a cloud database using satellite short burst data transmission. Though each independent sensor platform is a powerful sensing tool, the low cost permits large scale deployment. Hundreds or thousands of systems can be deployed over vast geographical areas and report to a remote database where cloud based machine learning algorithms can both develop global situational awareness and update local system algorithms based on learnings across the full network. Our first test of this system will be a set of ocean drifters to monitor sea state and enhance maritime awareness. This research was developed with funding from the Defense Advanced Research Projects Agency (DARPA). The views, opinions, and/or findings expressed are those of the author and should not be interpreted as representing the official views or policies of the Department of Defense or the U.S. Government.