Rosatom and Tsiprum: System for recording and storing dataset passports

2021: Establishment of a system for recording and storing dataset passports

Rosenergoatom Concern (part of the Electric Power Division of Rosatom State Corporation), CONSIST-OS JSC (a subsidiary of the Concern) and the Private Institution for Digitalization of the Nuclear Industry Tsiprum (Rosatom State Corporation) completed a pilot project to create an industry system for recording and storing dataset passports. This was announced on March 24, 2021 by Rostatom.

A dataset is a collection of data in terms of machine learning tasks and their description. The dataset passport contains information about its content, owner and purpose of use, and also allows you to evaluate its applicability to solving consumer problems, determine loading methods and options for subsequent use.

The project was implemented as part of the Rosatom program "End-to-end digital technologies and data management" and is aimed at creating a single platform for the industry register of datasets, machine learning models, methodologies for solving typical problems in the field of artificial intelligence.

The database has already loaded 12 pilot dataset passports created by Rosenergoatom and Zifrum as part of projects using artificial intelligence and the use of machine learning. The system undergoes the registration procedure in the Register of Russian Software.

Artificial intelligence and, in particular, machine learning are actively developing technologies in the industry. As of March 2021, a large amount of datasets have already accumulated, which are used to train artificial intelligence in various projects. In this regard, Rosenergoatom and the industry as a whole faced the issue of creating their register and realizing the possibility of reusing existing datasets in other projects. This will significantly reduce the time and labor costs of preparing data for creating models, "commented Oleg Shalnov, Director of the IT Project Management and Integration Department of Concern Rosenergoatom JSC.

Each dataset is placed in the registry along with a detailed description of its content, purpose and use history. This information allows you to evaluate the potential suitability of a particular set data for other tasks and its subsequent use. The presence of the registry also allows you to easily find the initial data on which this person trained, neuronet analyze and make the necessary adjustments to the model in case of malfunctions in systems with artificial intelligence.

In turn, Konstantin Kudashev, head of the Center for Digital Technologies of Rosenergoatom Concern, emphasized that the system created also solves the important problem of the safe use of artificial intelligence at industry enterprises.

{{quote 'The safety and effectiveness of artificial intelligence systems directly depends on the quality of the data on which machine learning models are built and trained. All our datasets are verified, tested on real models and working in industrial systems, which allows you to create more accurate models. Their very storage, located in our reference data center, ensures the safety, security and transparent use of all data sets, "said Konstantin Kudashev. }}

The creation of the dataset register is one of the first projects implemented by Zifrum in the direction of the development of digital technologies and a culture of data use in the nuclear industry.

The developed product allows you to track the use and usefulness of data, determine responsibility and take into account the contribution of people involved in the development of the field of artificial intelligence to the result of the development of the industry. The project demonstrated that when using digital technologies and combining the efforts of participants, data in the industry is a universal asset that can become a "fuel" for both existing and projected business processes, "said Anton Zapryagaev, deputy general director for end-to-end digital technologies and data management at Tsiprum.