Covering Disruptive Technology Powering Business in The Digital Age

The MUSKETEER project developed an industrial data platform for federated Machine Learning (ML) techniques—validated in healthcare and auto manufacturing—between 2019 until the end of 2021. By boosting data sharing across organisations, it opened up a very promising future for Artificial Intelligence (AI) in these and others potential sectors.

In two specific applications the federated and privacy-preserving AI ML was put to the test and proved to be a game changer.

• In the healthcare environment. The analysis of MRI scans to detect prostate cancer is becoming better, faster and MUSKETEER added a safety net in the training of AI models to improve image analysis processes.
• In the automotive industry. The classification of the welding point quality became significantly more effective and accurate, enabling new optimised maintenance processes.

The project showed that AI models can be trained in an entirely privacy-preserving way. The trainings were based on large amounts of data from different factories and hospitals and respecting EU regulations. The future use of similar AI support with these safety measures in place is very compelling for industry players in these two domains. The MUSKETEER platform enables the development of AI for improved diagnoses and faster processing in the area of medical imaging and car production—a big step forward for the health of our society and industry..

The MUSKETEER project would like to share its results publicly via an open-source software, https://github.com/Musketeer-H2020/, and a list of publications, dissemination material and public deliverables that can be accessed on https://musketeer.eu/publications/

“The MUSKETEER privacy-preserving federated machine learning solution enables a strong collaboration between different stakeholders in medical imaging to develop and improve AI models to support medical professionals in their decision making, improving their workflow and reducing reading time. As providers of advanced medical imaging solutions, we are excited to be able to use and expand a health instantiation of the MUSKETEER platform to keep collaborating with our partners and customers to deliver innovative AI tools to improve efficiencies and patients’ outcomes,” said Joao Correia, Innovation Manager at Biotronics3D, a medical imaging company.

“Using federated and collaborative machine learning techniques, Comau will be able to provide innovative maintenance services to their customers providing them more robust and more accurate predictive models, using data coming from different customers plants, while at the same time preserving privacy issues related to company data,” commented Massimo Ippolito, Head of Digital Innovation & Infrastructure at Comau, a robot manufacturer.

(0)(0)