Self-Managed Warehouse

Et Warehouse Control Systems (WCS) til å styre lageret, dette har typisk relativ simpel og regelbasert logik. Self-Managed Warehouse-prosjektet endrer dette med en forbedret tilnærming som bruker AI-teknikker som matematisk optimalisering, maskinlæring og digital tvilling-modellering. Prosjektet introduserer en toveis digital tvilling, en virtuell modell som speiler og samhandler med hver del av lageret i sanntid. Denne modellen muliggjør en sømløs integrering av vedlikeholds- og driftsplaner, og sikrer optimal effektivitet. Den utviklende digitale tvillingen dekker alle lagerkomponenter, fra lagringssystemer til robotarmer, og forbedrer og utvider seg kontinuerlig i omfang. Denne integrasjonen øker ikke bare effektiviteten, men forbedrer også vedlikehold og planlegging. I tillegg deles prosjektets funn og fremskritt med bransjepartnere, noe som fremmer utbredte forbedringer i lagerstyring. Dette initiativet markerer en overgang til lager som ikke bare er oppbevaringssteder, men smarte, selvstyrende økosystemer, som fører til en mer bærekraftig og effektiv fremtid.

An efficient Warehouse Control System (WCS) is essential for an efficient, sustainable, and labor-friendly warehouse. For too long, warehouses have been managed using myopic WCSs that lack proper integration with the physical components and processes within the warehouse, favoring short-term rewards versus long-term insights. The Self-Managed Warehouse project aims at a drastic improvement in warehouse operations and maintenance across the sector. Using an efficient suite of AI techniques composed of mathematical optimization, machine learning and digital twin modeling, a platform will be built that enables coordination of maintenance services with operational requirements, improves warehouse performance, and helps with scenario and what-if analyses of the system. Using up-to-date representations of the real world, decisions on one side of the table will lead to updates on the other side, ensuring that all involved parties are operating with updated best plans and at best possible efficiency. Through the project a two-way digital twin will be iteratively improved, covering a larger and larger portion of the warehouse – from storage systems to conveyor belts to robotic piece picking arms and beyond. At the same time, the decision support algorithms integrated with the two-way digital twin will be expanded to be able to cope with the new components, enabling an increasingly holistic improvement of the warehouse, able to take the performance of the entire system – and links to maintenance restrictions – into account. Better understanding of and support for maintenance and planning in warehouses will be good for the entire warehousing sector, and the project aims to disseminate and share its major findings with the industry through regular meetings and interactions with partners within and without the project consortium.