This project proposes an innovative Digital Information Management (DIM) concept, the AICHAIN solution, that aims at enabling the cyber-secured exploitation of large private data sets that belong to different stakeholders and that contain valuable information for ATM operations.
To overcome the stakeholders’ reluctance to share sensitive data, the exploitation will not be performed by exchanging the data itself but by articulating an advanced privacy-preserving federated learning architecture in which neither the training data nor the training model need to be exposed. This will be possible thanks to the innovative combination of two emerging DIM technologies: Federated Machine Learning (FedML) and Blockchain technologies.
The specific objectives of the research have been structured in three areas of research that will be covered with different levels of deepness:
The DIM technological solution.
The operational value of the DIM solution.
The governance & incentives aspects.
These are the specific objectives in these three areas:
OBJECTIVE #1: In the technological dimension, to define the AICHAIN Solution architecture as a potential SESAR technology enabler for the exploitation of private data value, and to implement a functional small-scale prototype for user validation and operational value experimentation.
OBJECTIVE #2: In the operational dimension, to demonstrate and quantify the operational value of the AICHAIN concept with an ATM use case in the area of A-DCB services.
OBJECTIVE #3: In the governance dimension, to develop an incentive mechanism that addresses the motivational aspects of the data owners in order to facilitate the adoption and the effective utilisation of the AICHAIN concept.
The concept developed in the AICHAIN project, illustrated below, will allow the secured and anonymised exploitation of valuable private datasets. It will combine two emerging technologies, federated machine learning and blockchain, to create a platform in which different stakeholders can provide inputs into federated machine learning models, while preserving privacy of business sensitive data. As a consequence, trained models will have access to datasets that would have never had otherwise. This project will target the development of said platform to the ATM community to improve predictability at network level.