New Insights from the ALCHIMIA Project Coordinator

Read below the latest update from the Coordinator of the ALCHIMIA Project, celebrating a year of groundbreaking work in digitally transforming Europe’s metallurgy industries through AI, sustainable practices, and green growth as we follow the exciting journey towards a smarter, more environmentally responsible metallurgical sector. 


The ALCHIMIA Project is currently celebrating a significant milestone in our groundbreaking project about the digitalisation of Europe’s metallurgy industries. Over the past year, our dedicated team has been tirelessly working to usher in a new era of industrial transformation, one that harnesses the power of artificial intelligence to energy-sustainable, green growth. 

 In our exploration, we take you on a journey through the remarkable progress we’ve made, showcasing the tangible impacts of our efforts on the European metallurgy sector. From revolutionising production processes to optimising resource management, our project is unlocking the full potential of AI, paving the way for smarter, more efficient, and environmentally responsible metallurgical practices. 

The main objective is to demonstrate how digitalisation and AI will propel the metallurgy industry into the 21st century while supporting the crucial goals of green growth and sustainability. 

ALCHIMIA has already completed its first year of the project. After a year of intense and exciting work, ALCHIMIA has reached its first milestones and enters the phase of implementing and integrating the technical solutions that will achieve the project objectives once the functional and non-functional requirements of the use cases have been identified. The main outcomes reached during this period are the first version of ALCHIMIA Federated Learning and Continual Learning framework, which provides features to allow training Machine Learning and Deep Learning models in a decentralised manner, including multiple capabilities in terms of aggregation and communication protocols. 

In parallel, based on initial data sets shared by CELSA, work is being carried out on the creation of local models, including analysis and cleaning of process data, adaptation of the dynamic Electric Arc Furnace (EAF) process model to the CELSA plant and initial adjustment of parameters; and the development of a demonstration web application with the main functions of scrap characterisation, optimisation of scrap mix and visualisation of the results of the EAF process modelling. The next step will be integrating the overall model into the Federated Learning framework. 

Over the last six months, project partners have successfully generated several deliverables documenting and detailing the aforementioned outcomes: requirements and human-centric recommendations, federated Learning infrastructure implementation and use-cases preparation. Deliverables will be shared soonhere ( 

On the 28th and 29th of September, partners gathered in Madrid to attend the 3rd General Assembly and present the current status quo of each of the work packages. The mid-term review is planned for the 26th of October, where the progress made will be presented to experts. 

ALCHIMIA is also glad to announce its participation in theESTAD conference in Düsseldorf, held in June 2023, and in theESTEP Dissemination Event in PISA, focused on the steel industry, held in March 2023. 


Thank you for being a part of this remarkable journey!