DeepSpace-AI – A Deep-Learning Based Offshore Monitoring System Using Satellite Imagery

Environmental monitoring and forecasting for marine, coastal and terrestrial environments plays a critical role in informed decision making for a wide range of stakeholders in government and regulatory agencies, private industry, and the scientific community. The need for high-quality rapid data and analytics is becoming increasingly important in the face of a changing climate, which threatens to disrupt ecosystems, alter global weather patterns, and increase the frequency and severity of extreme events across the globe. We propose to develop the DeepSpace-AI platform, for automated monitoring and forecasting of environmental phenomena on a global scale. The DeepSpace-AI platform will automate the import of satellite imagery and complementary data streams for areas of interest, and process this data using a range of machine learning models for object-detection, area-based segmentation, and forecasting. Additionally, the platform will include assisted annotation tools supporting efficient user-generation of custom analysis models.