LTU builds advanced AI factory for railway

AI technology that manages to predict the operational and maintenance needs of the railways in Sweden - and that facilitated the automation of maintenance and repair decisions. The AI Factory will simplify the decision-making process in maintenance through an advanced AI platform aimed for operation and maintenance of railway at Luleå University of Technology. Real-world experiments with a digitalised railway will be carried out throughout the country.

– We want to make the maintenance of the railway in Sweden faster, easier, cheaper, and more automated. Through the AI Factory, the railway players will have the opportunity to look into the future and make more precise planning and more efficient operation and maintenance of the entire railway system, Ramin Karim, says, project manager and Professor of Operations and Maintenance Engineering at Luleå University of Technology.

Automate the decision-making and maintenance

The basic idea of the AI Factory for railways is to automate the decision-making and execution of maintenance tasks for infrastructure, locomotives and wagons. This can be done through advanced computing technologies based on AI and an open secured sharing of algorithms and data. This advanced technologies and concepts will also be applicable in other industrial contexts. Through the AI Factory for railways at Luleå University of Technology, researchers want to accelerate the digitalisation of the railways to make rail transport in Sweden more efficient, effective, sustainable, safe, reliable and punctual. Real-world experiments with the AI Factory will be carried out all over the country - starting on Malmbanan and Haparanda railway.

-– We are not that far from being successful with this - and we will be the first in the world, Ramin Karim, says.

National initiative on data labs

The project, which has been granted over 8 million SEK, of which half from Vinnova, is part of a national initiative on eight new data labs that will lead to social and economic values through increased focus on data-driven decisions. The AI Factory will develop and provide a platform with AI services and AI components that can be composed by various railway players based on business and technical needs. This advanced technology will also be applicable in other industries.

– This investment is clearly a recognition of our research in Industrial AI which is built on research in eMaintenance and research in operation and maintenance of railway at Luleå University of Technology, Ramin Karim, says.

AI Factory's services and functions are open but secured

The AI Factory for railway at Luleå University of Technology consists of a set of smart services (cloud and edge-based). These enable various stakeholders in the railway sector such as authorities, industries, academia, and companies to get a better overview of the current and future capacity and permits of the railway - and thus optimize their operations. The AI Factory's services and functions will be provided through an open secured platform, which is a prerequisite for an industry-wide optimisation.

The project AI Factory for railway is done in collaboration with Alstom, Bombardier, the trade association Tågoperatörerna, Damill AB, Trafikverket, Infranord AB, Norrtåg AB, Omicold AB and Sweco Rail AB. The project has already received interest from other actors who will be involved in various contexts.

• https://www.mynewsdesk.com/se/lulea_tekniska_universitet/pressreleases/bygger-avancerad-ai-fabrik-foer-jaernvaeg-2955184?utm_campaign=send_list External link, opens in new window.
• https://www.jarnvagsnyheter.se/20191219/9681/ai-fabriken-digitaliserar-jarnvagen External link, opens in new window.
• https://www.infrastrukturnyheter.se/20200108/22740/ai-fabrik-jarnvag-automatiserar-drift-och-underhall External link, opens in new window.