Imagine a computer network that acts like a living organism, healing itself and organizing its own tasks without a human boss. Instead of one central brain making all the decisions, every small part of the network knows what's happening around it and adjusts instantly. It's like a swarm of bees working together to keep everything running smoothly, even when the system is overwhelmed.
Edge computing often suffers from instability under high loads and slow response times due to centralized management. Operators struggle with the complexity of managing thousands of microservices across diverse locations.
A decentralized AI management system featuring swarm-based orchestration, a local graph model for resource tracking, and an AI model app store.
If you are a factory operator dealing with lag in robot coordination — this project developed swarm-based orchestration that ensures stable performance even under extreme load. This reduces the risk of production halts caused by network complexity.
If you are a grid manager dealing with massive data flows from remote sensors — this project developed an edge-wide workload optimization service. This allows for faster local decision-making and reduces the environmental impact of data processing.
If you are a warehouse provider dealing with unstable microservices for fleet management — this project developed a cognitive cloud-edge system that safeguards stability. This ensures end-to-end transaction resiliency for your automated vehicles.
Based on available project data, specific pricing or implementation costs are not provided; however, the project aims to reduce the costs of cloud-edge management.
Yes, the project uses swarm technology and a decentralized graph model to manage resources across a wide edge-cloud infrastructure, moving away from centralized hub-and-spoke setups.
Based on available project data, licensing details are not specified, but the project includes an app store for sharing and rating the AI models used.
The project has already retrieved initial data from SCADA systems and network topology to validate its use cases.
The project period runs from 2023-01-01 to 2025-12-31, with validation currently happening in 3 specific scenarios.
The consortium is heavily weighted toward practical application, with an industry ratio of 45% comprising 5 companies, including 4 SMEs. With 11 partners across 9 countries, the project balances academic research from 4 universities and 2 research centers with direct industrial validation, suggesting a strong focus on commercial viability.
Contact INESC ID in Lisbon, Portugal
Contact us to explore the ACES AI model app store for your edge infrastructure.
