Abstract

We worked with a leading ocean intelligence company to develop a near real-time data ingestion and visualization platform for providing valuable insights over changing oceanography conditions. These insights are valuable to commercial fisheries, as well as agriculturists as it provides them with predictive analysis related to various species and changing ocean conditions.

About Our Client

Our client is a leading space technology and intelligence company, delivering services in Space Infrastructure and Earth Intelligence. We are working with them in the Ocean Intelligence space to help commercial fisheries optimize operations, monitor changing ocean conditions and protect aquaculture assets with near-real time data. The product suite is extendable to on-ground operations to help small and large agricultural farmers with prediction and communications capabilities.

Business Challenges

Business Needs

Near real-time analysis

Near real-time analytics to find the right type of fishes with optimum efficiencies.

Tracking weather conditions

Track changing weather, ocean, and climate conditions to maintain safe and efficient operations at sea.

Visualizing using proper tools

Monitor changing ocean conditions by global climate change as well as chemical contamination required proper visualization tools.

Challenges

Live analytics over data

Faster response to users, based on filters applied on the data.

Multiple output formats

Each layer was to be drawn differently, which required multiple output formats.

Domain knowledge

Every data layer needed understanding of the domain knowledge. Eg. how the direction of water currents change based on the water temperature.

Solution Details

1. Data Ingestion

   A data ingestion module is developed to fetch data from Satellite data feed. The data is stored as NetCDF and ESRI Shape files with segregation for date, region and products. Developed in Python as a daily cron job.

2. ETL pipeline

   An ETL pipeline is developed to read product files and transform it to create vector data – Shape files, GeoJSON layers and raster data – GeoTIFF layers.

3. Visualization

   A Powerful UI is developed, allowing users to manipulate, filter and sort the data for further analysis. MapBox GL JS is used to render geo-referenced images, contour lines and marker clusters.

Technology Used

Results and Impact

Revealing insights to optimize fishing operations

Near real-time data for fishing recommendations and ocean conditions helped the fishermen to locate the closest, most productive fishing grounds.

Monitor changing conditions

Various oceanographic data values like temperature, plankton helped researchers with the understanding of climate changes as well as effects of chemical contamination and its impact on marine life.

Expanding platform customer base

Feature based thin client deployment made it easy to reach a larger and wider customer base.

Conclusion

Developing a near real-time oceanographic data ingestion and visualization platform presents unique challenges, but with adequate research about the domain and experimenting with the technologies available, we were able to achieve our desired solutions. A map driven solution can further be extended for weather prediction, monitoring vessels in the ocean and for performing predictive fishing analysis.