Deliverables

The list of deliverables submitted to date is as follows:

Nr.TITLESUMMARY
D1.1Report on Use-case Definitions and Evaluation Metrics The main objective of the task underpinning this deliverable is to perform a high-level assessment and description of the use cases and show how they fulfil the main objectives of the SMART-PDM project and how they will advance the state of the art in its field.
D1.2Requirements for Use-cases’ Information ArchitectureThe main objective of the task underpinning this deliverable is to complete the high-level assessment and description of the use cases and determine the desirable information architecture to use at each case. For this purpose, this document aims at describing the functionalities that the monitoring system should comply with in the form of requirements; while presenting the different options for monitoring and signal management solutions to be used in the different use-cases that compose the SMART-PDM project. Proposed solutions are based on the needs of every use case and the discussion section present in this deliverable aims at detecting similarities, cooperation, and interoperation possibilities among all use-cases.
D1.3Privacy, Security and GDPRThis document describes the definition of privacy, security and General Data Protection Regulation (GDPR) related design and development considerations that will be carried out in SMART-PDM project. Following Deliverable 1.2, this deliverable focuses its work on how privacy and security will have to be handled on platform basis. This work also focuses on the implications that the new GDPR European legislation on the systems being developed, particularly on the handling of data Integration. It will initially be defined as per ten use cases for rolling mill, vehicles, wind turbines, seed drill/woodchipper, hydro power plants, milling machines, manufacturing laboratory, and home appliances. The home appliances use-case has specific considerations on privacy and ways to handle end-user specific data. All use cases have common cybersecurity considerations. In addition to GDPR, country specific ethics and privacy laws are also considered and elaborated in a separate section of this document.
D2.2Existing and desirable sensors to develop predictive maintenance for each use caseMonitoring of signals requires the identification of suitable sensors solutions. Sensor selection, deployment and characterization is therefore a fundamental step for the implementation of condition-based maintenance strategies and technical developments envisaged within the project.
Initially in chapter 2 this document details the signals that will be used to determine the condition of the production equipment components and the sensors systems currently available on the demonstrators.
In chapter 3, this document proposes a set of guidelines for selecting the sensor system solutions best suited to monitor the condition of use cases equipment. Such guidelines are intended to be applicable to different technological fields and will be tested in the context of the end users’ demonstrators.
D2.1Integration and Interconnectivity AnalysisThis document describes the definition of integration activities to be carried out in the SMART-PDM project. It includes brief descriptions of each integration definition, functional description, integration point description, requirements and constraints, activities and roles, risk analysis and schedule of integration activities. Integrations will initially be defined according to ten use cases for rolling mill, vehicles, wind turbines, seed drill / wood chipper, hydro power plants, milling machines, production laboratory and household appliances. The methods, tools and software used in the project are explained in the document. With the predictive maintenance method in the project, it is aimed to prevent a possible error that may occur with regular maintenance to the stands. In addition, saving time and costs is one of the main objectives of the project.
D2.3aNetwork and Storage Architecture ReportThis report presents the SMART-PDM network and storage architecture. The architecture facilitates data to be acquired from multiple sources, analytics to be provided by multiple vendors, and results to be communicated to all relevant parties. The architecture is based on standards and open source components.
Arrowhead framework provides tools for connecting data sources and data consumers by authenticating and authorising devices for communication between each other. This is done by Arrowhead managing a set of programmable rules on which device can access what devices within an Arrowhead cloud. Multiple Arrowhead clouds can also communicate with each other facilitating communication between devices managed by different Arrowhead servers. Small applications for setting up Arrowhead servers and utilizing Arrowhead are built as open source projects on GitHub under the VTT Operation and Maintenance organization (VTT Operation and Maintenance, 2021).
MIMOSA relational database model called CRIS (Common Relational Information Schema) is used in defining data structures that support maintenance operations. The implemented application requires only a subset of data structure definitions from MIMOSA. The data structure is used to store data such as measurement data and diagnostics results. The MIMOSA database model is run on open source MariaDB database server.
REST APIs are built for bi-directional data transfer between the devices and the MIMOSA databases. REST APIs provide an easy to use and secure interface to the database utilizing HTTPS protocol. An open source project Django REST framework is used to build the REST APIs.
D3.1Physical Signal and Attribute Analysis
related to each Use Case
SMART-PDM Task 3.1 deals with the third step of CRISP-DM, namely signal segmentation, outlier removal, feature extraction, clustering, feature comparison and knowledge extraction. This way, it will be possible to identify whether the captured signals are enough to identify the deterioration, to estimate the remaining useful life of subsystem and systems. During the process, it may also be necessary to define new variables by combining different input signals using different mathematical models, which can better serve the purposes of model predictions that take place in later tasks.
The deliverable will attempt to give insight into the types of signals and data attributes that will be pre-processed under this task for the purpose of modelling needed as part of Tasks 3.2 and 4.2.
D3.2aImplemented Advanced Data Analytics for Diagnostics, Prognostics and Advisory generationManufacturing accounts for 15.96% of EU GDP (World Bank), is undergoing immense yet gradual Industry 4.0 transformation with the help of advancements including predictive maintenance, whose market size is estimated to be worth USD 4.9b by 2021 (Markets and Markets).
SMART-PDM’s objective is to provide diagnosis and prognosis information while rendering the underlying technology financially feasible. The project improves Overall Equipment Effectiveness (OEE) of the manufacturing industry by, e.g., implementing cost-efficient Predictive Maintenance (PDM) utilizing low-cost sensors.
One of the key techniques in PDM is data science, i.e., analysis and modelling of measurement data. When combined with domain knowledge, e.g., understanding of possible failure mechanisms, automatic and reliable diagnostics, prognostics, and decision support can be provided. After describing the data analysis techniques used in the different use cases of the project in Deliverable 3.1, this document aims at describing the generalities of advanced data analytics related to diagnosis, prognosis, and advisory generation, to latter describe the different actions taken in the different use cases of the SMART-PDM project.
D3.2bData Analytics SoftwareTask 3.2 is primarily regarding the identification of the techniques required to:
Diagnose the current system’s status: this can be based on prior knowledge of the system, by experts.
Forecast the evolution of the system in order to determine future maintenance points.
Prescribe new working conditions for the best performance, at the same time, attending asset management policies.
While D3.2a focuses on the results obtained from the implementation of such techniques, D3.2b, being a software deliverable, focusses on the software that is used to obtain such results. Similar to D3.2a being distinguished from D4.2a in that industrial results are only included in D4.2a and not D3.2a, D3.2b is intended to emphasise parts of the software relating to the overall predictive and prescriptive techniques whereas D4.2b places effort to include software features relevant to industrial applications.
D4.1End users’ requirements related to FMECA analysisIn this report Failure Mode Effects and Criticalities Analysis (FMECA) and Criticality Matrix (CM) for each of the use cases of the project are included. The aim of carrying out FMECA and CM analysis has been to identify the failure modes, their occurrence rates, severities and effects that will constitute the base for the development of an effective maintenance management. This will be crucial for the development of the predictive maintenance policies / rules that will be derived as the result of this project.
In most of the uses cases, there already exists knowledge about the systems, machines, or processes that will be analysed as use-cases. This is a valuable source of information for predictive maintenance tools development, since it allows pointing out the most relevant parts, issues and possible solutions (from the maintenance point of view). This knowledge has been crucial to apply FMECA and CM methodologies.
For some use-case owners, this has been the first time they have implemented FMECA and CM methodologies. All partners have found these tools to be effective to extract relevant information about the functioning of their systems that serve to identify most relevant failures, set priorities and response actions. In fact, these tools have served, to structure the available knowledge about the systems of each use case.
D4.2aReport containing the implementation of predictive maintenance techniques – first prototypeThe aim of the report is to describe the methods of predictive maintenance techniques that have been implemented in various pilots. The report presents each use case as a whole, including methods for operation and predictive maintenance, which will enable the machine to be operated better and, on the other hand, provide an estimation of the Remaining Useful Life (RUL) of the selected components. In other words, this report describes the first prototype on a case-by-case basis to which preventive maintenance has been applied.
D4.2bSoftware for the Implementation of Predictive Maintenance Techniques – First PrototypeThis document describes the system requirements for and the features of the use-case D4.2b as well as the description of the graphical user interface and the results of the first prototype.
D5.1Dissemination PlanThe dissemination activities of the SMART-PDM project are part of Task 5.3 (Dissemination). The primary objective for dissemination is to create strong awareness of the SMART-PDM project and technologies both on the European but also on the national level, with the aim to multiply its impact and subsequent exploitation chances. The planned standardisation efforts will foster a further industrialisation of the SMART-PDM results and create a solid base for investment and commitment of the SMART-PDM partners. Commercialisation will be supported by giving SMART-PDM results – from today’s perspective the developed techniques around data analytics. The proper set of dissemination activities will be carefully studied, starting with the identification of all the different stakeholders that may be interested in the main outcomes of the SMART-PDM project and to raise their individual interest.
D5.2Preliminary Exploitation PlanSMART-PDM’s objective is to acquire manufacturing data to provide diagnosis and prognosis information while rendering the underlying technology financially feasible.
In order to achieve this goal, we created a very complementary and conducive consortium with 11 use case demonstrators ambitious to see the trial results, 9 integration & tools providers and 9 technology and knowledge providers all waiting to test out and further improve their technologies, all with a balanced breakdown of 10 SMEs, 10 Large Enterprises and 6 RTOs which are spread across 5 countries.