DataSpaces - Using the Marketplace
Future Software Technologies
Semester programme:Open Learning - Main
Research group:Interaction Design
Project group members:Benjamin Takaki
Juul Hoezen
Georgi Fidanov
Dimitar Kostenarov
Project description
Our group implemented and configured the Marketplace within a Data Space architecture, utilizing reference models such as IDS, Gaia-X, and FIWARE.
The Marketplace served as the central facilitating and orchestrating layer. We applied several use cases to address all elements of a Data Space comprehensively, focusing on both the processes and technology involved.
Context
This project sits within the broader research domain of Data Spaces and digital infrastructure, specifically the marketplace and access-control layer that lets organisations discover, negotiate, and consume each other's data and services in a trustworthy, decentralised way.
The applied use case models a business-to-business setting in which one organisation offers metered energy reports and managed Kubernetes (cloud infrastructure) services, and a second organisation discovers, orders, and consumes them, with all access mediated by verifiable credentials and machine-readable usage policies rather than static contracts.
We implemented this scenario using the FIWARE Data Space Connector (FDSC), following the DSBA reference architecture and aligning with IDS, Gaia-X, and FIWARE specifications. A second project group addressed the identical use case using the Eclipse Dataspace Components (EDC) connector instead, allowing both groups' results to be compared as two independent reference implementations of the same data space requirements.
This comparative angle is part of the project's broader value: it shows how the same conceptual marketplace and trust requirements can be satisfied by different open-source connector stacks, and where implementation effort and maturity diverge.
Results
The project's primary product is a fully deployed Minimum Viable Dataspace built on the FIWARE Data Space Connector, running on a local Kubernetes (k3s) cluster, covering the complete lifecycle a real data space participant would go through: trust anchor registration, verifiable credential issuance, ODRL-based access policy enforcement, marketplace product catalogue and ordering through TMForum APIs, automated contract management, and post-purchase resource provisioning (issuing operator credentials and granting access to a managed Kubernetes cluster only after a completed purchase). Every phase, including negative cases such as credential or token rejection for unauthorised roles, was verified to behave correctly.
On top of this deployment, we built an interactive dashboard as the main supporting deliverable. Rather than requiring a reader to manually copy-paste curl commands from a markdown guide, the dashboard walks the user through every step of the local deployment guide interactively: it executes the underlying commands directly, displays live status checks for each service and credential exchange, and visually confirms whether a step succeeded or failed before allowing progression to the next one. This turns a script- and terminal-heavy procedure into a guided, self-validating walkthrough, reusable both as a teaching tool and as a regression-style check after redeploying the environment.