Within the first reporting period, most of the activities focused on the development of methodology and software for the various algorithmic components of the RECLAIM solution. A key step regarding the methodology was a literature review of lifetime extension strategies for Circular Economy and the definition of their specifications and impact on Key Performance indicators. A key component it th Decision Support Framework (DSF). As regards the development of algorithms for the DSF, the „Reliability Analysis Tool“ currently supports the statistical analysis of failure events, the Failure Mode Effect Analysis and the Reliability Block Diagram analysis, relying on failure events data collected. The „Machinery Operational Profiling“ provided the mathematical definitions of the three main profiles of the industrial equipment to be studied in RECLAIM: Health, Production and Performance. For the „Fault Diagnosis and Predictive Maintenance Simulation Engine using Digital Twin“ component, the first prototype of the Digital Twin was developed, and algorithms for fault diagnosis and predictive maintenance were developed and integrated with a draft version of the „Optimization Toolkit - Refurbishment & Remanufacturing Planning“ component, for which an optimization planning algorithm was developed. Regarding the „Prognostic & Health Management Toolkit“, degradation models were defined for particular pilots, and an algorithm finding stop correlations was developed, while hardware and software interfaces for data acquisition were also tested. Another activity was the definition of the methodology and the development of the initial algorithm for the „Cost Modelling & Financial Analysis Toolkit“, based on the activity based costing approach. On top of the previous components, the „DSF Core“ was developed in a draft version as a component providing integrated decision support for refurbishment, remanufacturing and other defined strategies. Apart from the above components, which are part of the DSF, other algorithmic components were also partly developed. These are the „augmented reality (AR) mechanisms“, for which the indoor localization and 3D registration, AR visualization and feature recognition modules were developed within this period, as well as the „In-situ Repair Data Analytics“ component, for which machine learning image processing models were developed for optical control of product quality. Most of all above algorithms were implemented using artificial data based on the pilot requirements.
The second period concentrated on the integration of RECLAIM technology into the Pilots. Firstly, interfaces have been implemented to make production data available for further processing and evaluation of the algorithms developed. To do so, various sensors have been mechanically and electrically integrated and also interfaced to the data communication mechanisms. Secondly, various RECLAIM solutions have been integrated. Where neccessary, those solution have been tailored to the specific need of the respective Pilot. Among others, the following RECLAIM solutions are now available at Pilot #1 Gorenje: Predictive maintenance algorithms for spot and seam welding, Pilot #2 Fluchos: Degradation monitoring of shoe making machine, Pilot #3 Podium: Digital twin model for shop floor simulation for impact evaluation of planned and predictive maintenance activities, Pilot #4: Harms&Wende: AR glasses for In-Situ Repair and faster failure finding and repair for friction welding machines, and Plilot #5 Zorluteks: An Adaptive Smart Sensorial Network and Digital Retrofitting Infrastructure for machine inspection and quality assurance of fabric's bleaching processes.
