Global Environment
In the face of a changing climate, which threatens to disrupt ecosystems, alter global weather patterns and increase the frequency and severity of extreme weather events, the requirement for high quality rapid data analytics is more important than ever.
Synthetik leverages machine learning and modeling and simulation tools to enable real-time monitoring of environmental phenomena.
DeepSpace-AI
Real-time environmental monitoring and forecasting
Overview
Synthetik’s DeepSpace-AI leverages deep learning with large amounts of earth observation data for monitoring and forecasting a wide range of environmental phenomena via a web-based platform. The platform allows users to set regions of interest to automate the import of freely available satellite image data e.g., Landsat and Sentinel satellites and commercial data. The downloaded imagery is processed using machine learning models to identify objects of interest.
01. Objective
Develop a web-based software platform capable of automated identification of environmental phenomena. Synthetik chose to demonstrate the feasibility of this technology using the identification of harmful algae blooms (HAB).
02. Process
Environmental monitoring is controlled by firstly defining an ‘event’ and then training a machine to identify events from satellite imagery. The success of the platform relies on the successful integration of data import at volume and powerful analysis modules powered by machine learning.
03. Results
The technology provides a single-source clearinghouse for satellite data ingestion for the development of efficient machine learning models. Heatmaps of algal concentrations were generated using the DeepSpace-AI platform to prove viability of the project to the U.S. National Oceanic and Atmospheric Administration (NOAA). The scope to expand the platform is limitless.
DeepSeaVision-AI
Computer vision and artificial intelligence platform shapes marine environmental management
Overview
Synthetik’s DeepSeaVision-AI is an integrated hardware and software platform for automated real-time monitoring of marine wildlife. The system alerts users to the detection of marine wildlife within a user-specified area to enable the prevention of entanglement events between wildlife and offshore infrastructure.
01. Objective
Develop a web-based software platform capable of automated monitoring of marine environmental phenomena
02. Process
As part of a feasibility demonstration, Synthetik trained computer vision models on classifying whales and cetaceans (0.5-1m per pixel) and using visual and audio data collected from deployed camera and hydrophones, perform machine learning inferencing at the edge to detect the marine mammals, as well as birds and vessels.
03. Results
The DeepSeaVision-AI platform was developed through a grant awarded by the U.S. National Oceanic and Atmospheric Administration (NOAA) and is currently undergoing a pilot at the University of New Hampshire. The platform has also won additional funding to evaluate a potential site for offshore finfish aquaculture facility operated by Manna Fish Farms.
Custom solutions to fit your needs
Not seeing what you’re looking for? We create solutions for a variety of needs in insurance, global environment, energy, and defense & security. Reach out and we’ll be in touch.