Self-describing Digital Assets and Their Applications in an Integrated Science and Engineering Ecosystem

An integrated science and engineering ecosystem requires the sharing of digital assets across many different participating organizations. Digital assets include data sets, AI models, system configuration, papers and technical reports etc. that are exchanged across different organizations. Due to a large diversity in the syntax and semantics of the digital assets, their use across and within organizations is fraught with difficulties. If the digital assets were self-describing, their usage and exploitation in organizations different than the ones producing them would be much simpler. However, the addition of self-description needs to be done in a light-weight and flexible manner in order to leverage the existing digital ecosystem, with the appropriate trade-off between extensibility, scalability and security. In an open-source collaborative research venture in an effort called the Enterprise NeuroSystems Group, several companies and universities are working together to create a light-weight self-description mechanism and a catalog to facilitate exchange of self-describing digital assets to support a variety of use-cases. In this paper, we describe the approach for adding self-description to existing digital assets, the architecture of this catalog, the fundamental design choices made for the approach and the architecture of the catalog, and the use-cases for collaborative science that can be enabled using such a catalog.