One-stop-shop for Artificial Intelligence and Machine Learning for Official Statistics Project (AIML4OS) is an ESSnet collaborative project involving 16 countries, started on 1 April 2024 and will last 4 years.

The overall aim is to develop a ‘One Stop Shop’ for AI and ML in official statistics; that is, a platform which provides a single entry point for ESS staff to access a coherent set of capabilities for implementing AI/ML based solutions. This will provide them with tailored guidance and assistance in deploying AI/ML solutions within adequate methodological and implementation frameworks.  The project is developed through 13 WPs, 6 supporting implementation and 7 dedicated to use cases. 

The knowledge and the framework realised and use cases experiences developed by the Consortium will allow them to use AI/ML to enable the development of more efficient and effective ways to deliver growing demand on official statistics.

Expected outcomes from this project include:

• the delivery of a framework for developing AI/ML solutions to be used in the context of official and European statistics
• the provision of access for ESS staff and partners to established and proven AI/ML solutions/resources to be leveraged in the context of official statistics production
• the encouragement of engagement from ESS organisations with AI/ML for innovation purposes and to facilitate their understanding and realisation of the benefits of AI/ML
• the delivery of economies of scale and resources through cooperation inside and outside the ESS and the acceleration from ideas regarding AI/ML to actual production.

Key activities for this project are to:

• Develop, maintain and evolve a coherent set of relevant capabilities – including methodologies, guidelines, sandboxes, labelled data, processes, methodological, implementation and quality frameworks for implementing AI/ML based solutions in official statistics across the ESS
• Set up a platform/hub providing a single entry point for ESS staff to access relevant capabilities
• Provide support and guidance for the integration and maintenance of relevant AI/ML based solutions in ESS organisations through training and active and efficient support
• Build communities around open-source solutions developed and maintained by ESS members
• Share ideas, experiences, success stories and lessons learned to stimulate innovation based on the use of AI/ML 
• Enable and facilitate the transition from development and experimentation of AI/ML based solutions to actual production.

The project will develop seven use cases to test and promote AI/ML-based solutions for the production of official statistics. 

The aim is to enhance the creation of official statistics by integrating advanced artificial intelligence and machine learning techniques related to these use cases. Additionally, it includes a comprehensive review of the best practices developed to date within National Statistical Institutes (NSIs). 

The use cases were selected to ensure broad applicability across the EU. Four of the use cases aim to build on areas where AI/ML is currently being considered and, in some cases, actively used (WP7 AI/ML on Earth observation data, WP8 and WP9 focusing on Editing and Imputation in official statistics by AI/ML, and WP10 exploring the potentials of AI/ML for Classifying and Coding). These use cases will assess the progress made to date in these areas before considering ways to enhance robustness, deepen development, and broaden use across Member States. Additionally, two completely new areas for AI/ML use cases will be explored. The first new area is linked to the use of Large Language Models, while the second one investigates the feasibility of mapping supply chain networks starting from company-level data.

The project is carried out by the following partners and is financed by the European Commission. 

To stay informed about the latest developments of the project, please subscribe to the newsletter: essnet-aiml4os@istat.it.