Industry 4.0 is really a revolution in all aspects, but maybe not for all companies and maybe not just now. There are many definitions of Industry 4.0 and most of them include the use of AI and digitalization. One of the ideas with AI is that we can automate decision making based on perceptions of the surrounding environment. Perceptions of the surrounding environment can also be called data and data quality will become an issue worth exploring. Not all companies seem to give sufficient attention to the problem with poor quality data and the annually associated with poor data quality is estimated to more than 600 billion dollars.

Data quality can be defined in the four areas of Intrinsic data quality, Contextual data quality, Representational data quality, and Accessibility data quality. The dimensions of each area say more about the areas. Intrinsic data quality is about Believability, Reputation, Accuracy and Objectivity. This area is the essential properties of data quality. Contextual data quality contains the dimensions of Value, Relevancy, Timeliness, Completeness, and Appropriate amount of data. This area contains elements of missing or incomplete data and is heavily dependent on the context. The third area, Representational data quality, contains the dimensions of Interpretability, Ease of understanding, Representational consistency, and Concise representation and has to do with how the are formatted. Last Accessibility data quality concerns Accessibility and Security. All these areas and dimensions gives a coherent picture of data quality.

The lack of data quality is causing problems for companies looking to implement AI in different levels of decision making. If AI is implemented based on poor quality data, the results of the AI decisions will also be poor. This undermines the whole idea with Industry 4.0 and the use of AI. At Linköping University, we currently research this issue with poor quality data with the aim to be able to do simulations, digital twins and eventually AI implementations. But before all that, we are looking into methods to increase the quality of existing data and to be able to collect new high-quality data. A recent application to the Swedish funding authorities brings together three companies, a research institute, and Linköping University. The results can make Industry 4.0 an endeavor for all companies in the near future.

Prof. Janerik Lundquist

Linköping University