Summary
- Any complex manufacturing system generates large amounts of data, and this data is often underutilised and little value is generated. Due to FA3D2’s highly flexible nature, it is an extremely complex system with many sources of variability which require understanding and compensating for.
- The integrated digital twin system of FA3D2 simplifies the gathering and contextualising of data via an OPC UA-based service bus. This data is collected and aggregated in a cloud-based environment for analysis.
- Machine learning approached and other analytical tools are being developed to analyse the data, update virtual twins or test proposed changes in the digital twin, and then deploy these changes to the shop floor via the same digital pipeline as the virtual commissioning process.
- Key to this approach is integrating the skills and knowledge of the workers on the shop floor, who work in conjunction with the AI approaches to achieve hybrid decision making – the combination of human and machine intelligence.
Results type(s)
Unfold all
/
Fold all
More information & hyperlinks
| Country: | GB |
| Address: | Advanced Manufacturing Building, 522 Derby Rd, Lenton, Nottingham NG8 1BB |
Geographical location(s)
Structured mapping
Unfold all
/
Fold all
Demonstrators, pilots, prototypes
Significant innovations and lessons learned, training aspects
Technologies and enablers
Comment:
Python / Tensorflow / Pytorch – Used for the creation of bespoke machine learning algorithms and analysis processes.
ICT performance characteristics
Pathways
Comment:
Siemens Mindsphere will be used to collect and analyse data from the shop floor; cloud-based.
Amazon Web Services – Currently used to host cloud data and machine learning algorithms.
(European) Digital Innovation Hubs (EDIH)
Search
Contacts
Not specified (see website if available) or see associated project
Associated products/services
Grafana – Used for the creation of bespoke data visualisation solutions.