Vi använder kakor (cookies) på vår webbplats för att den ska fungera på ett bra sätt för dig.
Genom att surfa vidare godkänner du att vi använder kakor. Vad är kakor?
This work package will assure the overall project activities to be planned, executed and followed up according to the goals and visions, i.e. the overall scientific and co-production process. The work package includes activities for expectation and coordination management, communication of both research and practical results. The co-production companies have representatives in the project group.
Well performing machines are important to ensure high quality and to avoid unscheduled or unnecessary production stops. Production stops due to failing equipment decreases the productivity and might hurt product quality and increase costs.
Maintenance relies heavily on monitoring of the production process and collection of the related data.
The technological advances made during the last decades within the Internet of Things (IoT), Big Data, Machine Learning (ML), and Artificial Intelligence (AI) have made it possible to continuously monitor and analyze thousands of machine parameters. These technologies create new possibilities for improving the maintenance process. In general, Machine Learning and Artificial Intelligence have a significant role in Industry 4.0. ML and AI represent new requirements to the process of data collection in order for the entire system of predictive maintenance to function correctly. In particular, the data must conform to the principles of Validity, Accuracy, Completeness, Consistency, and Uniformity. In the AHIL project we will leverage these Industry 4.0 technologies by prototyping a predictive maintenance application that satisfies the data collection and processing requirements as well as improves productivity of the manufacturing process.
In this work package we focus on collecting, storing and analyzing machine data using big data and machine learning technologies. An intricate part of WP2 is to analyze the machine data to create useful insights into the behavior of the machines and to predict future machine performance. The insights and predictions will then be disseminated to the users through WP4 apps or, if applicable, automatic action will be taken. From previous and on-going research, we know that data-driven analysis techniques can be used to predict future machine performance including, but not limited to, predictive maintenance applications. One of the partner companies has an, in house developed, digital platform which collects over 168 000 sensor values each second from a set of 21 machines. All this data comes from just a small part of the companies’ production plants. In WP2 data- and analytic-driven scenarios will be designed, implemented and brought into production. As an example, one scenario is to use vibration measurements to determine CNC spindle health. In the first stages the analysis will be performed offline but as the solutions mature the goal is to disseminate the findings to the users in real time, e.g. via apps, which will be handled by WP4
The research in the WP3 investigate challenges and opportunities for initiation or extended applications of production flow simulation i.e., discrete event simulation (DES) as a potential digital twin, and thus, as a tool for decision support in the era of industry 4.0 manufacturing. We began the research withing WP3 with a case study, where we initiated the application of DES at a company (with “fairly low” digitalization emphasis) to whom this technology was novel. Throughout the project we are further extending the implementation of DES in co-production with the companies to delve into the areas of real time updates and application of predictive algorithms (e.g. could be artificial intelligence (AI) or machine learning (ML)) coupled with the DES digital twin. Running in parallel with the DES study that focuses on the quantitative data collection and technical aspects of production flow simulation is an interview and focus group study. The AHIL project focuses on artificial and human intelligence therefore we stress the importance of studying both simulated data and human understanding and viewpoints. The interviews/focus groups are performed with personal from all levels at the partner company i.e., management group, economics, human resources, manufacturing engineers, logistic and purchase personal, as well as operators and production personal. Until now we have completed 13 formal interview or focus groups, and further are planned over the course of the project. Additional data is collected through meetings, workshops, and seminars in co-production with the industrial partner companies. The research is extended by coupling Master thesis work to it. Those students are employed by the company during the course of their thesis work and are placed at the company site. This close collaboration and interaction with the project partner companies on many levels’ emphasis the aspects of co-production and work integrated learning within the AHIL project. A particular strength of the research within WP3 is the combination of DES modelling (mainly quantitative data) with interviews and focus groups with company employees on many levels (mainly qualitative data). The interdisciplinary aspect of the research facilitates and emphasizes an increased understanding of both artificial and human interaction and learning (I-WIL) when understanding how to design and application of technologies within Industry 4.0.
The complexity of introducing discrete event simulation as a new tool for decision making is highlighted, where we emphasize the combination of technical and human knowledge and involvement yet necessary to understand and to draw conclusions from the collected and simulated data. There may be value in reflecting of the implication that human preconceptions may have on the results, denoting that the data analysis and DES models can be valuable when striving towards objectivity. The results also demonstrate that the data analysis has given valuable insights into production characteristics, otherwise hard to pinpoint, and this information can accordingly be addressed at the partner companies. Thus, the results reveal opportunities for how the initiative of introducing discrete event simulation as an anew tool in the wake of industry 4.0, can act as a catalyst for improved decision making in future manufacturing.
This area of research is about investigating and understanding the digitization of industrial work linked to the fact that more and more machines in production are connected with sensor technology that can measure real machine capacity, which is part of what is often called Industry 4.0. By implementing and connecting so-called IoT systems (the Internet of Things), it is possible to measure and analyze machines and the real capacity of production systems. That is, how much machine capacity is actually used. However, this means that large amounts of data are generated, so-called Big data. The advantage is that directly connected machines create opportunities for fast and direct action for and communication between operators, technicians and managers for delegated decision-making at several levels. How data is to be structured and visualized both locally in a computer and on screens in a workshop then needs to be designed based on several different user needs, ie based on different needs and functions in the industrial companies' direct production. But it entails a number of challenges for creating the benefit of generated data and how qualified analyzes are performed. What type of qualitative decisions can be made based on real-time data and how does it affect the business' planning of operator and technician work that has consequences for production planning? It also has consequences for how operators and technicians are to integrate new data and knowledge in relation to already proven and experience-based working methods and routines. All in all, this affects the engineering profession in the form of roles and competencies and how to integrate new learning into direct industrial work.
The digitalization of AI-empowered manufacturing industry will have long-term effects on organizational strategic work and thus also affect how individuals - and the organization as a whole - choose to benefit and experience the transformation. Learning processes within the organization can be threatened by low overall company support and lack of competence to handle new disruptive technology in production work. In this WP we will identify and understand the consequences and challenges for the industrialization and scaling up of AI initiatives. We will develop a strategic work-integrated learning model on how AI- and data-driven applications (such as those described in WP2-4) can be incorporated into an organizational learning approach in the context I4.0.