Distributed, Secure and Energy-Aware AI Infrastructure for Industrial Applications
| Reference number | |
| Coordinator | Ericsson AB |
| Funding from Vinnova | SEK 12 712 966 |
| Project duration | June 2026 - June 2028 |
| Status | Ongoing |
| Venture | Advanced digitalization - Industrial needs-driven innovation |
| Call | Shared, sustainable and secure AI infrastructure through advanced digitalization |
Purpose and goal
Develop and validate: • A modular, pod-based architecture, including performance and scalability testing. • Secure orchestration for multiple tenants, measuring security, isolation, and performance trade-offs. • Energy-efficient orchestration for distributed AI, measuring energy savings and peak power reduction.
Expected effects and result
Expected effects • Greater access to computing resources for industrial AI. • Reduced reliance on cloud providers outside Europe. • More secure, energy-efficient AI deployment in industry Expected results • A distributed AI infrastructure connecting modular compute pods across multiple sites while keeping data under local control. • Methods for running GPU-intensive AI across sites. • Secure multi-tenant operation • Scheduling methods that account for power limits and improve energy efficiency.
Planned approach and implementation
• The project will build representative, scaled-down AI infrastructure pods and connect them to a testbed for controlled experiments across distributed, heterogeneous environments. • A digital twin of compute, network, and energy behavior will be created, calibrating it with physical measurements to predict large-scale systems. • The performance and scalability will be validated across pods, multi-tenant security, isolation, and energy efficiency under realistic workloads and grid conditions.