Safety interlock recommendation system

1 . A safety interlock recommendation system, comprising:

at least one process data source,

an edge device,

wherein the process data source is configured for providing process stream data to the edge device;

wherein the process stream data includes real-time operational data;

wherein the edge device comprises an operational technology edge application unit, OT edge application unit, and a stream analysis unit;

wherein the OT edge application unit is configured for providing operation technology stream data, OT stream data;

wherein the OT stream data comprises safety-relevant data from the OT application unit;

wherein the stream analysis unit comprises an online machine learning model, being configured for determining online analysis data using stream data comprising the provided process stream data and the provided OT stream data;

wherein the OT edge application unit is configured for determining a short-term recommendation for proactive actions including adjusting control parameters of components in the process data source to avoid potential safety interlock events using the online analysis data;

wherein the online machine learning model is configured for executing an online explorative analysis to determine the online analysis data;

wherein the online explorative analysis comprises association mining and root cause analysis;

wherein association mining comprises detecting operational patterns in the stream data; and

wherein the root cause analysis comprises detecting a root cause in the stream data for a potential safety interlock event.

2 . The system of claim 1 , wherein the stream data comprises dynamic time-series data directly relating to a controlled process of the industrial plant.

3 . The system of claim 1 , further comprising a cloud platform, wherein the edge device is configured for determining batch data using stored stream data; wherein the cloud platform comprises a batch analysis unit and a cloud application unit; wherein the batch analysis unit comprises an offline machine learning model configured for determining offline analysis data using the batch data; wherein the cloud application unit is configured for determining a long-term operational recommendation using the offline analysis data.

4 . The system of claim 3 , wherein the offline machine learning model executes an offline explorative analysis to determine the offline analysis data; wherein the offline explorative analysis comprises association mining and root cause analysis.

5 . The system of claim 3 , wherein association mining comprises detecting operational patterns in the batch data and wherein the root cause analysis comprises detecting a root cause in the batch data for a potential safety interlock event.

6 . The system of claim 3 , wherein the cloud platform comprises cloud storage, wherein the cloud storage is configured for providing storage data to the batch analysis unit; wherein the storage data comprises additional stored data relating to the safety interlock recommendation system; and wherein the offline machine learning model is configured for determining the offline analysis data using the provided storage data.

7 . The system of claim 3 , wherein the cloud application unit is configured for determining a re-engineering recommendation using the offline analysis data; wherein the re-engineering recommendation comprises potential design enhancement for an existing interlocking logic.

8 . The system of claim 3 , wherein the cloud application unit is configured for determining the long-term engineering recommendation and/or the long-time operational recommendation using the provided storage data.

9 . The system of claim 3 , wherein the cloud storage is configured for receiving offline user feedback on a former long-term operational recommendation and/or former re-engineering recommendation; and wherein the offline machine learning model is configured to be retrained using the offline user feedback.

10 . The system of claim 1 , wherein the edge device is configured for receiving online user feedback on a former short-term recommendation; and wherein the online machine learning model is configured to be retrained using the online user feedback.

11 . The system of claim 1 , wherein the online machine learning model and/or the offline machine learning model uses Bayesian Networks.

12 . A method for recommending a safety interlock, comprising:

providing, by a process data source, process stream data to an edge device, wherein the process stream data includes real-time operational data;

wherein the edge device comprises an operational technology edge application unit, OT edge application unit, and a stream analysis unit;

providing, by the OT edge application unit operation technology stream data, OT stream data, wherein the OT stream data comprises safety-relevant data from the OT application unit;

determining, by an online machine learning model of the stream analysis unit, online analysis data using stream data comprising the provided process stream data and the provided OT stream data by executing an explorative analysis, thereby determining online analysis data;

determining a short-term recommendation for proactive actions including adjusting control parameters of components in the process data source to avoid potential safety interlock events using the online analysis data;

wherein determining by the online machine learning model online analysis data comprises executing an online explorative analysis to determine the online analysis data;

wherein the online explorative analysis comprises association mining and root cause analysis;

wherein association mining comprises detecting operational patterns in the stream data;

wherein the short-term recommendation provides proactive recommendations for a user to drive away from the undesired process downtime effect;

and

wherein the root cause analysis comprises detecting a root cause in the stream data for a potential safety interlock event.

13 . The method of claim 12 , wherein the stream data comprises dynamic time-series data directly relating to a controlled process of the industrial plant.

14 . The method of claim 12 , further comprising a cloud platform, wherein the edge device is configured for determining batch data using stored stream data; wherein the cloud platform comprises a batch analysis unit and a cloud application unit; wherein the batch analysis unit comprises an offline machine learning model configured for determining offline analysis data using the batch data; wherein the cloud application unit is configured for determining a long-term operational recommendation using the offline analysis data.