210527 Data Week Data-Logistics-for-AI-900-506

Fairs & Events

May 27, 2021

BDVA Data Week 2021 | Workshop: Data Logistics for AI

Data sharing is already a challenge but data sharing for AI brings about some additional complexities, like enabling the training of models while maintaining data sovereignty. Because of the complexity, this is more than data sharing so we call it “data logistics”. Data logistics for AI breaks down functionalities in manageable applications easing the challenge of responsible data sharing for AI.

The workshop aims to chart the challenges of data logistics for AI, to present some innovative developments to address those challenges, and to identify how the solution is firmly are embedded in European Data Spaces.

For example, Giulia Giussani from the International Data Spaces Association will explain the IDS standard for data sovereignty and why sovereign data sharing is essential in data logistics. Ruslan Bernijazov from Fraunhofer IEM will present the ambitious AI Marketplace, a data marketplace that brings together producers, service providers, and researchers to work together to revolutionize product creation with artificial intelligence. The AI marketplace is a use case for the IDS standard.

The workshop does not want to be complete in its definition of “Data Logistics for AI”, instead, the audience is invited to contribute to this first chart of the territory. The approach presented has emerged from the Dutch program on AI (NL AI Coalitie), but since AI cannot be applied in a meaningful way without international collaboration, it needs to be synchronized with the European community.

Register here: www.big-data-value.eu/dw21-registration

Date/Time
Date(s): May 27, 2021
9:30 am - 11:00 am

ADD EVENT TO YOUR CALENDAR

See all Fairs & Events

Stay updated with us

Subscribe to our newsletter

MONDRAGON ASSEMBLY EGOKIA

MONDRAGON ASSEMBLY - an international firm specializing in the development of automatization and assembly solutions for various sectors. It operates 6 production plants around the world and is part of the largest cooperative group in the world | IKERLAN Technical research center and member of IDSA

CHALLENGE:

EGOKIA solution allows collaborative learning and knowledge sharing between geographically distributed plants, integrating headquarters in Basque Country and domain experts. Thanks to the data space, data, and AI assets can be securely shared throughout different corporate sites and a new knowledge-based economy emerges.

SUCCESS:

Deploys an architecture based on IDSA connectors, linking Mondragon Assembly’s data and AI platforms from different corporate sites. With this new reliable data space, we extend the lifecycle of artificial intelligence models, enhance their capabilities, and facilitate trusted access and use of its assets by EGOKIA participants.

BENEFITS:

NEW BUSINESS MODEL: A new business model based on AI services and a knowledge-based economy. RELIABLE AI SOLUTIONS: A "standardized", flexible, and secure solution to break data silos to share and complete AI assets and processes between trusted organizations and production plants. Measurable benefits can be listed as: Up to 30% quality on models thanks to the application of federated learning techniques, reduce inversion and time to market on IA-based solutions or lower communications loads. STANDARIZED DATA SPACE – The organization is ready to take part in the knowledge economy with other entities, domain experts, and technology providers.

COMPONENTS:

Data Space connectors 8.0.2 Metada Broker 5.0.3 Identity Provider 1.6.0

Increased visibility of order delivery process

VTT | Wärtsilä | Uwira

CHALLENGE:

The use case revolves around expediting and enhancing transparency between a focal company of a manufacturing ecosystem (OEM) and its supplier in their production order management processes. The existing communication methods have proven inadequate in facilitating efficient and adaptable exchanges between the two entities. By facilitating transparency and fostering a more seamless integration between the IT systems of both companies, the management of product orders can be streamlined, resulting in heightened efficiency and a reduction in errors.

SUCCESS:

The objective is to facilitate a reliable and seamless flow of data between the OEM and its supplier by establishing a dedicated data space. The developed solution will be seamlessly integrated with pre-existing IT infrastructures of both companies and the solution is grounded in the framework of the IDS reference architecture.

The OEM can selectively disclose data from their production environment to the supplier without the fear of data leakages. Respectively, the supplier receives only the data that are essential for them and there is no need to locally extract meaningful insights and valuable information from big data sets.

BENEFITS:

The established data space solution has the potential to significantly enhance transparency between the OEM and the supplier. This gives rise to several potential benefits, including:
Effective Communication: Transparent communication helps in conveying accurate information about product specifications, requirements, changes, and timelines.
Risk Management: Sharing information about potential risks, such as changes in demand or design modifications, allows suppliers to proactively address these issues. Furthermore, in times of disruptions, such as supply shortages or unexpected market shifts, transparent communication helps both parties respond more effectively.
Cost Efficiency: Transparent communication allows both parties to work together to find cost-effective solutions.

COMPONENTS:

IDS connectors - serve as the communication interface linking the two companies and their existing IT systems. With IDS connectors, a data provider can define the rules and conditions (usage policies) under which data are shared with a data consumer.
DAPS - enables connectors to authenticate themselves using X.509 certificates. After successful authentication, the DAPS issues OAuth2 access tokens for connectors, who need these tokens to access the services and data of other connectors.
Metadata Broker— stores information about the data end-points offered by the participants. The broker provides a query interface for connectors and can hence be considered as the search engine in this use case.

LEARN MORE:

Show Use Case