Imagine having thousands of small computers and sensors spread across a city, all talking to a central cloud. Instead of a human trying to manually tune every single device, this system acts like a smart autopilot that manages itself. It learns on the fly and adjusts settings to save energy and keep things running fast without needing constant manual updates.
Managing thousands of diverse IoT and edge devices manually is impossible at scale. Companies struggle with high energy costs, network delays, and the complexity of deploying software across different hardware types.
An AI-driven management system using a hierarchical agent architecture and portable containers for automated resource control.
If you are a city operator dealing with massive data from traffic and utility sensors — this project developed an agent-based AI system that automates resource management. This ensures low latency and energy efficiency across the city's digital network.
If you are a farming tech provider dealing with remote IoT devices in fields with poor connectivity — this project developed a system that manages data processing close to the source. This reduces the need for constant cloud connection and optimizes battery life of field devices.
If you are a network provider dealing with the complexity of managing heterogeneous hardware across a wide area — this project developed a way to deploy applications using portable containers. This allows for flexible and isolated execution across different types of servers.
Based on available project data, specific pricing for the end-product is not listed, as the project is funded by a EUR 5,711,250 EU contribution for research and development.
Yes, the project uses realistic system simulators for scale-out experiments and tests the system using two real-world application-specific testbeds in smart cities and agriculture.
Based on available project data, the project emphasizes openness and extensibility through an API for pluggable ML models, though specific licensing terms are not detailed.
The system is designed to interface with popular control mechanisms and uses portable container-based technology to ensure it works across different types of infrastructure.
The project period runs from 2023-01-01 to 2026-01-31, suggesting that final validated results will be available by early 2026.
The consortium is highly balanced for commercialization, featuring a 50% industry ratio with 6 industrial partners, including 4 SMEs. With 12 partners across 8 countries, the project combines academic research from 4 universities and 2 research centers with practical industrial application, increasing the likelihood of market adoption.
Contact PANEPISTIMIO THESSALIAS in Greece
Contact us to explore licensing opportunities for the agent-based AI architecture.
