Visualisation training at Kellogg’s

The installation and introduction to PI Vision enabled many of the staff to access their data in ways that were not possible before, enabling data driven understanding and decisions to be made.

|

SHARE

Background

At Kellogg’s, the main insight users had to their production data, was through a series of pre-created PI ProcessBook screens in a shared folder.  These screens were managed by a core number ospecialists to control any version changes and make sure the right data was mapped.  This resulted in a core number of screens that were frequently used; however, users were constrained to the data that was available on these screens. 

Requests for more data was possible, however the process was lengthy given the limited number of people with the insights required to search for PI Tags and technical knowhow to build up the PI ProcessBook screens.  As such, users made do with the ProcessBook screens that already existed, and only requested additional data when absolutely necessary. 

Solution

As a part of the PI Vision installation, the PI ProcessBook screens were uploaded, and an existing PI AF database model was registered so it may easily be queried by users.  This AF model had limited development time dedicated to it however, as the subject matter specialist also had to field requests for the PI ProcessBook screens and tag data requests.  As such, most users were unaware of this database’s existence, or how to query AF for real time data. 

Following the PI Vision installation, a short workshop was conducted to show the users through the PI Vision technology, also working through the existing PI AF model.  While this model was in its infancy, this already had a lot of the data required for most users.  

Working through PI Vision showed users how they could access their already familiar screens, as well as methods to quickly and easily build up dashboards for any of their specific requirements.  By using AF, users learnt how to navigate the AF database by exposing it in PI Vision, as well as how to quickly extract the relevant information needed for their analysis in any dashboards they wish to create.  This ultimately allowed them to intuitively find the data that they needed for their analysis, eliminating the need to learn specific tag conventions or requests for individual tags. 

As users are being shifted to use PI Vision instead of PI ProcessBook for their data needs, time is freed up for the subject matter specialist to work on the AF model, and add tags that may be requested. 

Other PI AF features were then able to be used within PI Vision, such as element relative dashboards, where a common asset type model in AF results in a common dashboard definition in PI Vision for all similar assets. 

Conclusion

Users can now build up screens to their exact requirements, sharing them with others, knowing that the underlying data can be trusted and where exactly it has been sourced.  This includes any ad-hoc screens for a quick analyses, or investigations for rapid-value realisation.  This was not possible before, due to the time and effort required by multiple parties to source the required PI tags and construct the needed screen.

Picture of Archie Gale
Archie Gale

Head of Innovation

By introducing the operational team to PI Vision and PI AF, we were able to open the door to what was already possible with Kellogg's data they already had.

Andrew Todd

Chief Executive Officer at GTS Group

Other case studies