APIs to Support AI in Environmental Research
Artificial Intelligence (AI) and Machine Learning (ML) workflows are increasingly used to undertake data science analysis to answer environmental questions. This proposal focuses on creating and utilising data-driven APIs that power workflow pipelines and thereby make data driven AI/ML easier to achieve. Such software technology will enrich the NERC Environmental Data Service (EDS) ultimately gain greater insights from this data to support societal decisions. Data-driven APIs are fundamental to the provision of an AI-powered integrated Digital Research Infrastructures. This proposal aims to develop consistent approaches to API design, development and deployment across EDS, sharing expertise and experience. This enhances both the efficiency and consistency of API provision and will enable integrated access to cross-discipline scientific data to meet existing and new use cases from collaborators. It will widen engagement in cross discipline proposals, thereby upskilling across environmental sciences. It will enable new multi-disciplinary scientific analysis and lower the barriers to entry. The proposal will look beyond the EDS to ensure approaches are consistent with other initiatives/standards across the UK and internationally.
The project will enable:
- Better Data. All EDS data centres need to provide accurate, accessible and interoperable machine-readable data. This underpins the ‘strong foundations’ of environmental data for the UK.
- Better sharing of data to ensure better access to data across and beyond the environmental sector and through facilitating automation of these capabilities, and through enabling interoperability.
- Better analysis of the data. We need to lower the technical barriers to environmental data analysis, preventing the need for additional, often expensive, resources or specialist skills.