Process analytics with the viadee Process Warehouse
Once you have taken first steps in process management and business process automation, you may want to analyze and optimize your running processes. With the viadee Process Warehouse we created an innovative Open Source solution for this purpose.
Process management und optimization of automated business processes
The BPM lifecycle describes four fundamental phases of business process management – modeling, execution, analysis and optimization of processes. Modern process engines enable automated process execution and provide analysts with a very useful set of execution log data. While broad tool support is already available for modeling and executing processes, there are currently no solutions for a process-oriented analysis of the log data. Thus, potentially valuable data remains unused and potentials for optimization are not leveraged. The viadee Process Warehouse wants to close this gap.
Drawbacks of traditional data analysis concepts
While executing business processes, process engines collect detailed event logs. These logs contain information about what is actually happening inside process instances and therefore provide a very useful data set for process analysts. Processing the collected log data with traditional data warehouse systems seems like an obvious answer. However, these systems are not able to interpret BPMN models which means that a direct reference to the process is lost. Consequently, metrics have to be defined on a low level of abstraction (e.g. based on process element IDs). This approach requires analysts to manually identify critical parts of the process model and adapt ETL processes accordingly for new process versions. Not only does this require a high maintenance effort, but potentials for optimization remain undiscovered.
viadee’s approach: Analysis in the context of the process model
The viadee Process Warehouse (vPW) allows the definition and visualization of process metrics directly on the process model. In combination with intelligent filters (e.g. on process elements, process variables or time spans) the vPW enables analysts to answer complex process-related questions and reduces the aforementioned maintenance effort. The idea behind the vPW is to provide an analytical BI perspective on process log data by processing the data in near real time. This implies that process metrics don’t have to be defined in advance in the process-modeling phase. Instead, an explorative and intuitive analysis in the third phase of the BPM lifecycle is possible. The metrics are visualized directly on the process model so that the BPM lifecycle can be completed without losing the context of the process in the analysis phase.
Key goals of the viadee Process Warehouse
- High-level definition of process metrics
- Visualizations directly on the process-model diagram
- Analytical BI perspective on processes
- Computation of metrics in near real time
- Process Intelligence: control of live metrics on running process instances
- Integration with the Enterprise Data Warehouse
Try viadee Process Warehouse
In our demo environment, you can try out the viadee Process Warehouse (vPW) with an example process first hand. The process is running in a Camunda process engine and continuously generates new process data. The vPW queries these data and stores them in an Elasticsearch index. This allows you to visualize process metrics in the web frontend of the vPW as well as in Kibana. The following figures illustrate the architecture as well as some basic features of the vPW. Moreover, the following two video tutorials help you with the first steps.
Our demo environment uses Camunda as Process Engine: Here you find the Camunda Tasklist The task list allows to start the demo process and execute user tasks. You use the same user and password as for the vPW. Kibana allows to visualize process performance indicators in terms of diagrams: Here you find Kibana (User: kibana, Password: kibana)