Process Assurance Runtime

The  component is an end-user application for process quality assurance. At the core it enables the user to monitor and analyse the manufacturing process. Users can manage prediction and optimisation components, KPIs and alarms in separate collections (known as recipes). Alarms can be set which are triggered if failures are about to happen or simply notifications can be configured which are sent if certain criteria are met. If sufficient data is provided the PA Run-time component can even suggest specific action that should be taken to prevent future defects. To achieve these goals the PA Run-time uses internally different Prediction and Optimisation Run-time (PO Run-time) components each solving one specific data processing problem. This component interacts with several components in ZDMP platform, such as PO Run-time, Monitoring and Alerting and AI analytics Run-time.


  • Integration of process monitoring and control into ZDMP platform

  • Management of prediction and optimisation components, KPIs and alarms in separate collections (known as recipes)

  • Ability to learn from the expert actions

  • Application Run time providing Minizdmp platform that allows to deploy PO Run-time component locally.

Play Video

Interface to manage PO Run-times

The user can view all available versions of the PO Run-time component that are saved in a registry. Each version can be deployed which creates a new instance of the specific PO Run-time version. Once an instance is deployed it can be configured, started, or stopped. For this the PA Run-time component provides a flexible UI that assists the user where possible. For example, the configuration UI automatically queries the parameters that need to be set from the running PO Run-time instance and displays them on the UI.

Real time process monitoring

Users can set-up KPIs to keep track of any data provided by the message-bus, including quality predictions or raw sensor data. A KPI is automatically configured for any message-bus data that is used as an input or output by a running PO Run-time instance. All configured KPIs can be inspected using the Grafana UI.

Query process optimisations or predictions

The Process Quality Assurance Run-time keeps track of all running PO Run-time components the user has connected. For every Prediction and Optimisation Run-time the user can trigger a onetime computation at any point. The result is, independent of the original output configuration, displayed in the PA Run-time UI.

Self-learning of control actions

An instance of a PO Run-time component is specialized to learn to control a specific device by observing the process values and the control actions that are taken by an expert. The computational layer of the PO Run-time consists of an incremental learning algorithm that learns to map the process values to actions that are taken by operator. This PO Run-time is shipped with Process Assurance Run-time component; however, it could be modified and re-packaged using PO designer.

Additional resources

Learn more about i4FS by visting the project website for general information, the wiki for information about the core components, the Technical Manual for API documentation, and downloading the repository’s source code.

Training Academy

Get a better understanding of the global architecture and information flow.

Source code

Our source code is opensource and available on our Gitlab repository.

Software Documentation

Read our easy to follow documentation to learn how to use the i4 Components.

Software Tutorials

Follow our step by step tutorials to create your first zApp.