CeADAR leads EU consortium on energy-efficient AI

19 Feb 2024

Minister Dara Calleary, TD, and Dr Ricardo Simon Carbajo. Image: Angela Halpin

One of the objectives of the MANOLO project is to create a handbook with practical guidelines that will help companies comply with the upcoming AI Act.

CeADAR, Ireland’s Centre for Applied AI, is leading an €8.7m EU research project that seeks to make AI systems more trustworthy and energy efficient.

Known as MANOLO, the Horizon Europe-funded project brings together 18 partner organisations from eight EU countries with the aim of building tools and algorithms that will help reduce the amount of energy, memory and storage required to run AI models.

It will also be responsible for creating a framework to reverse a growing trend of declining user confidence in the technology.

Dr Ricardo Simon Carbajo, director of innovation and development at CeADAR and project coordinator of MANOLO, said that the aim is to create novel algorithms to make AI more trustworthy and energy efficient.

“The solution will plug into current AI systems, which will help organisations in Ireland and Europe achieve their energy efficiency targets,” he said.

“It will also provide tools and frameworks to evaluate and improve the trustworthiness of their AI systems while empowering humans and contributing to the good use of AI.”

One of the objectives of the MANOLO project is to create a handbook with practical guidelines on how to apply the framework and tools. This can help companies conform to the EU AI Act, which has been in development since 2021 and was passed by the EU Parliament last June.

Over the next three years, researchers plan to develop and test their AI tools before rolling them out for validation in a variety of industry settings.

Some of these tools, for example, will be embedded in humanoid robots operating in manufacturing and healthcare environments, in low-power sensor devices for brain-computer interface – such as those that monitor sleep – and in mobile phones.

The resources will also be tested in neuromorphic chips, computer processors modelled on the neural systems of the human brain which can perform tasks faster than standard computer processers while using less power.

Minister of State of the Department of Enterprise, Trade and Employment, Dara Calleary, TD, said that CeADAR’s leadership of MANOLO demonstrates Ireland’s ability to “punch above its weight” in terms of innovation in AI and machine learning.

“It is hugely important that we make AI systems more energy efficient given that use of the technology is likely to become even more widespread in the coming years,” said Calleary, who launched the project at University College Dublin (UCD) recently.

“We also welcome the work to reverse declining user confidence in AI and to provide more clarity around how AI systems work. These tools are going to underpin the workings of our society going forward and it’s vital that the public has total confidence that those in charge of their development are seen to be taking a fair, equitable and humancentric approach.”

Find out how emerging tech trends are transforming tomorrow with our new podcast, Future Human: The Series. Listen now on Spotify, on Apple or wherever you get your podcasts.

Vish Gain is a journalist with Silicon Republic

editorial@siliconrepublic.com