A real-time stream storage and analysis platform for underwater acoustic monitoring

  • Authors:
  • J. P. Hayes;H. R. Kolar;A. Akhriev;M. G. Barry;M. E. Purcell;E. P. McKeown

  • Affiliations:
  • IBM Research Division, Dublin 15, Ireland;IBM Research Division, Yorktown Heights, NY;IBM Research Division, IBM Technology, Dublin 15, Ireland;IBM Research Division, IBM Technology, Dublin 15, Ireland;IBM Research Division, IBM Technology, Dublin 15, Ireland;Biospheric Engineering Ltd., Barna, Co., Galway, Ireland

  • Venue:
  • IBM Journal of Research and Development
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

We describe a distributed, real-time system for the collection and analysis of underwater acoustic data. The system uses a number of preprocessing steps to classify and detect acoustic events and to identify and compensate for gaps in the data stream. Different event-detection techniques are applied in a distributed manner on the incoming data stream from each sensor to aid in the indexing and storage of the data. Other event-detection techniques process multiple simultaneous streams to identify and classify events of interest. Building upon the deployed system, a stream analytical platform provides data handling, preprocessing, and analytics in real time. These analytics identify and classify anthropogenic, environmental, and animal noise (a significant amount of which occurs outside the audible range of human hearing) and ascertain the direction of the noise source.