File:2020-03-03 IBM PAIRS Presentation Siyuan Lu IBM Watson Research Center.pdf

IBM PAIRS - Big Geospatio-temporal Data and Analytics as-a-Service Siyuan Lu, Manager and Principal Research Staff Member IBM T. J. Watson Research Center, Yorktown Heights, NY 10598, USA

  The rapid growth of geospatial data from satellites, drones, weather models, IoT sensors opens unprecedented opportunities. However, the size and complexity of such data also poise challenges for conventional geographic information system (GIS) tools. To fully exploit their value and leverage AI, required is a new paradigm of platforms with the following salient features: (i) scalability - handling petabytes of data, (ii) harmonization - masking the complexity of metadata, projection etc., (iii) complex query - searching of data at a “pixel level”. (iv) “in-data” computation – avoiding data movement through the internet. In this context, we present IBM PAIRS (https://ibmpairs.mybluemix.net) – a next generation geospatio-temporal platform designed around distributed key-value stores and architected to provide the features noted above and high-performance data ingestion, query, and analytics. Exemplary applications in agricultural prediction and climate change assessment will be discussed to highlight how contextual insights can be rapidly gained through a variety of “cross-layer” queries and analytics to reveal relationships.