Project idea and potential
It is argued that the digital transformation of the manufacturing industry will increase the focus on the factory design and maintenance phases. In other words, efforts and job opportunities are expected to shift from deployment and operation phases to the design, monitoring and maintenance of smart and highly automated production systems. This shift will demand a digitally literate and highly skilled workforce in the manufacturing industry. This virtual testbed project specifically targets these two phases. The main purpose is to enable competitive and sustainable production systems by using virtual tools for factory design and advanced data analytics (machine learning) for smart maintenance.
Current OEE (Overall Equipment Effectiveness) figures are 50-55% on average in the manufacturing industry. This level has basically stayed unchanged for several decades. A positive trend is necessary to realise the full benefits of industrial digitalization. Future smart production systems need to operate on levels well above current world-class targets (85%). The design and maintenance phases are identified as key enablers for this change.
Brief background and state of the art
Advances in technology and management practices have removed many barriers to improve production systems’ performance, but many still remain unchanged—such as short-termism, lack of responsiveness and flexibility, and local sub-optimization. The current trends in automation, digitalization, data analytics and machine learning hold strong potential in addressing these shortcomings by improving data management, integration of sustainability considerations and strategic focus in planning activities.
Decisions for designing production systems requires a solid base incorporating complex, multi-dimensional aspects. The physical position of production resources needs to be defined with accurate spatial data, based on real environments but analyzed offline in virtual models. 3D laser scanning is a well-established technology for capturing spatial data of real environments. The scan data are used to create realistic models of e.g. facilities and shop floor layout as point clouds. These environments can potentially be over layered with several different layers of information to aid decision-making, which will be demonstrated in SUMMIT.
Data-driven decision support tools in the maintenance of future digitalized production systems is highly prioritized in industry. The application of machine learning techniques is emerging and the current focus is on the use of more extensive data sets and more predictive mindset compared to traditional condition monitoring. The combination of different data sources in the same analysis is also an increasing trend. These areas are explored by manufacturing companies world-wide. Cross-disciplinary work procedure is necessary to create decision support tools with the data interpretation to train the algorithms effectively. A closely linked challenge is to collect and manage high-quality data, which is necessary to reach the predictive and prescriptive levels of data analytics.
SUMMIT aims at increasing sustainability, efficiency, and robustness of production systems by utilizing the full potential of production data analysis in the design and maintenance phases.
The project goals focus on generating pilot cases in the design of digitalized factories and Smart Maintenance using machine learning for prescriptive analytics. The project will also analyze sustainability aspects of the developed approaches and provide dissemination materials such as learning modules for professional and university education. These projects goals will lead to following effects in Swedish industry:
Long-term effect goals (10-20 years)
• Resource efficient production systems applying eco-efficiency and circularity principles with zero-impact or net-positive effects on the environment
• Zero-failure production systems and OEE above 90%
Mid-term effect goals (3-10 years)
• Socially sustainable and attractive work-places
• Highly competitive Swedish companies with OEE above 85%
• Higher automation levels enabled
• Higher operational efficiency and reduced environmental impact
• Prescriptive maintenance solutions developed and used in industry
The collaborative testbed environment used in SUMMIT is the Virtual Development Laboratory (VDL) at Chalmers. VDL is an existing arena for collaboration between industry, academia, and institutes. The VDL will provide remote access to the testbed through internet as well as local onsite access in order to interact, showcase and evaluate the digitalization benefits developed. Cross-disciplinary teams of industry experts, data scientists, and production domain experts collaborate on developing and testing data-driven decision support tools. Project partners will provide software functionality to develop and test innovative solutions related to data analytics and virtual tools.
- Sigma IT Consulting (Private, Sweden)
- Saab (Private, Sweden)
- Scania CV AB (Private, Sweden)
- SSAB AB (Private, Sweden)
- Göteborg Energi AB (Private, Sweden)
- Microsoft (Private, Sweden)
- Combitech (Private, Sweden)
- China-Euro Vehicle Technology (CEVT) AB (Private, Sweden)
- Royal Institute of Technology (KTH) (Academic, Sweden)
- Siemens (Private, Sweden)
- Preem (Private, Sweden)
- ATS AB (Private, Sweden)
- University of Skövde (Academic, Sweden)
- University of Skövde (Publisher, Sweden)
- Volvo Cars (Private, Sweden)
- Fraunhofer-Chalmers Centre (Research Institute, Sweden)