RAMEN – Resilient Industrial Platform for the Advanced Visualisation of Predictive Maintenance, is an EIT-Manufacturing project, aims to bring a new software solution for condition monitoring and predictive analytics to the market. The focus relies on a systematic approach, combining the relevant AI algorithms, concepts, and specific solution into an industrial ecosystem.
This project will deliver a novel distributed and resilient platform for intelligence processing of production and process data with a focus on predictive maintenance and augmented reality for maintenance, based on micro services and federated machine learning. Additionally, containerisation technologies will provide an abstraction layer enabling scalability, resilience and flexibility.
The RAMEN solution will be applied and validated in a white goods use case.
The RAMEN Platform is a cloud-based resilient platform to support manufacturers in their maintenance activities and operational strategies. The RAMEN Platform offers its added value by combining the strengths of the results of two H2020 FoF projects, UPTIME and SERENA. It augments UPTIME’s predictive analytics with SERENA’s advanced & intuitive AR/VR visualization technology. RAMEN creates a flexible, resilient and scalable platform with unique ease-of-use and applicability to many different use cases.
The RAMEN Platform supports the integration and orchestration, management and replacement of multiple components on the basis of docker technologies and enables predictive maintenance in an industrial environment. It allows bridging the gaps between the shop floor operators, data science and engineering, as well as lowers the entry barrier for companies including SMEs to apply predictive maintenance solution with its containerized micro-service approach.
RAMEN will integrate and mature pre-commercial hard and software developed in the UPTIME project for predictive maintenance into the platform. The goal is to deliver fully mature Predictive Maintenance component which can be easily configured and operated by stakeholders in the relevant manufacturing processes.
The Predictive Maintenance component will encompass the entire data value chain from data acquisition condition monitoring and predictive maintenance. It will contain an intuitive calculation flow designer with an extensible library of algorithms and AI methods.
The Advanced Visualisation component will be integrated into the RAMEN platform as a service for interaction with maintenance-related information and the results of the Predictive Maintenance Component to shopfloor operators. Work in SERENA will serve as a basis introducing the key elements for creating the visualization component, supporting AR display.
The component will provide a user-friendly interface and support near real-time data streaming and asset visualization, superimposing maintenance-related information to the asset visualisation as well as predictive analytics.
Predictive maintenance can help manufacturers to avoid costly critical failure. Through combination of intuitive visual programming component for machine learning, analytics and prediction with the Augmented Reality, RAMEN will make predictive maintenance intuitive and easy to understand for different stakeholders in the manufacturing processes.
RAMEN solution is aimed to:
• improve customers’ OEE between 5% and 20%
• reduce MTTR by 10% – 20%
• increase MTBF by 20% – 30%
• reduce TCM by 15% – 25%
• extend component life by 10% – 15%
Use Case Definition
WP 1, led by Whirlpool, deals with the definition of the industrial use case by the industrial end users, the detailed functional and technical requirements deriving from the described use case along with the definition of the validation criteria of RAMEN platform.
WP 2, led by COMAU, deals with the design and development of the project cloud-based platform as well as the integration of the predictive maintenance and visualization modules to it along with edge components, i.e. gateways and legacy systems.
WP 3, led by BIBA, deals with defining, maturing & deploying the predictive maintenance component, with maturing the component especially with regards to user interaction and integrating the solution of the previous task. It also improves and finetunes existing solutions in predictive maintenance.
WP 4, led by LMS, deals with the target on the development and adaptation of a visualization service supporting augmented reality display, for providing assistance to the maintenance personnel. The provided information shall include the machine status, its evolution in time, as well as supported key performance indicators.
BIBA – Bremer Institut für Produktion und Logistik GmbH is an engineering research institute developing technical and organizational solutions for production and logistics in close cooperation with industry.
Laboratory for Manufacturing Systems and Automation® of the Department of Mechanical Engineering and Aeronautics, in the University of Patras, Greece, is oriented on research and development in cutting edge scientific and technological fields.
COMAU is the worldwide leader in manufacturing flexible, automation systems and integrating products, processes and services that increase efficiency while lowering overall costs.
Whirlpool Corporation is the world’s leading global manufacturer and marketer of major home appliances. In Europe, Whirlpool EMEA SpA has a sales presence in over 30 countries and an extremely articulated footprint.