IEEE International Conference on Emerging Technologies and Factory Automation (ETFA),
September 2024 · Padova, Italy · doi: 10.1109/ETFA61755.2024.10710791
The potential benefits of digital transformation for manufacturing companies include reduced costs, increased interconnectedness, and improved adaptability. Semantic Web technologies such as IRIs, RDF graphs, OWL ontologies, and SPARQL requests are a well-known and actively researched approach for supporting these transformation efforts. One challenge with this concept of knowledge augmentation is identifying where and how to integrate such semantic technologies into a manufacturing system, as it could require frequent translations into other non-semantic representations, which may entail a loss of expressivity and other disadvantages. Therefore, this work aims to use semantic technologies in a knowledge-augmented robotic manufacturing platform as directly and natively as possible. This approach includes the semantic modeling of manufacturing processes (similarly to flow charts) and context knowledge such as generalized mechanisms of how to apply them. All of this semantic knowledge is instantiated and persistently stored in a Robot Knowledge Base application, which implements mechanisms to automatically derive the next robot skill invocations and their parameter values during process execution. These semantic description models and the Robot Knowledge Base were tested in simulation as well as integrated into a physical mobile robot system with an articulated arm tackling an industrial use case.