Imagine trying to run a giant brain inside a tiny smartwatch; it usually doesn't fit and drains the battery instantly. This work creates a way to shrink those digital brains so they fit on small devices without losing their intelligence. It also helps decide whether a task should be handled by the device itself or sent to a powerful remote server to save energy.
AI models are becoming too large and energy-hungry to run on edge devices, leading to high costs, environmental harm, and a lack of user trust in automated systems.
A toolset of algorithms for model compression, meta-learning, and dynamic task allocation. It includes a data management system for tracking assets and a benchmark system for evaluating AI performance.
If you are a medical device manufacturer dealing with high power consumption in wearable monitors — this project developed a stack of algorithms that makes AI models lighter and more efficient. This allows real-time health monitoring on the device without draining the battery.
If you are an industrial robotics firm dealing with lag and high data costs from cloud processing — this project developed hardware-aware AI and neuromorphic chip support. This enables robots to process complex decisions locally and instantly.
If you are an IoT infrastructure provider dealing with the massive energy footprint of thousands of connected sensors — this project developed dynamic allocation tools for the cloud-edge continuum. This reduces the carbon footprint of your network operations.
Based on available project data, there is no specific pricing or cost structure mentioned as the project is currently in the research and development phase.
Yes, the project is designed for the cloud-edge continuum and will be validated in verticals like health, manufacturing, and telecommunications to ensure industrial viability.
Based on available project data, specific licensing terms are not provided, but the project aims to deliver a toolset for AI practitioners.
Yes, the project uses the Z-Inspection methodology for TrustworthyAI assessment to help systems conform to the new AI Act regulation.
MANOLO is designed to integrate with ongoing EU projects developing the next operating system for the cloud-edge continuum and the AI-on-demand platform.
The consortium is well-balanced for commercialization, featuring 19 partners across 8 countries. With a 37% industry ratio (7 industrial partners, including 7 SMEs), there is a strong link between academic research (7 universities, 3 research centers) and market application, ensuring the tools are built for real-world business needs.
Contact University College Dublin for technical specifications on the AI toolset.
Contact us to connect with the MANOLO consortium for early access to the toolset.
