New feature: Celonis Process Simulation in pycelonis

Celonis Process Simulation (in pycelonis)

Simulate process changes before implementing them.

With the release of version 1.5, we are making our Process Simulation capabilities available in pycelonis. This allows users with python knowledge to use our simulation capabilities inside the ML Workbench. Read more about pycelonis 1.5 here.

If you know what simulation can do and want to start implementing it, jump to How to use Process Simulation in python of this post directly (last part of this post).

We are also going to release a fully integrated, easy to use simulation experience in Celonis Studio, which is planned for the end of Q1 2021.

Introduction to Process Simulation

Modern organizations are looking for a data-driven way to know beforehand how big or small changes in their processes could impact the process outcome.

This is due to the fact that process changes have various challenges

  • High strategic risk which cannot be calculated in advance
  • What-if scenarios hard to estimate and evaluate
  • Costly optimization efforts with no way of measuring their impact

Celonis Process Simulation allows you to simulate the full impact of changes in your processes and lets you evaluate all your optimization scenarios.

Benefits

  • Simulate different options for resource reallocation and process adjustments
  • Get a great sense of where and how to take action based on impact and outcome
  • Get a dynamic visualization of all possible scenarios beforehand
  • Calculate an adaptive simulation model incorporating day-to-day changes

How it works

  1. Extract a Digital Twin by leveraging the actual process data: This creates a statistical representation of the process, including information about which resources work on which activities, their working calendar, and where in the process work piles up and creates bottlenecks.

  2. Create simulation scenario and run simulation: To see how this change impacts the overall performance and behavior of the process, the user runs a simulation scenario. Our integrated discrete event simulator calculates a simulated process model based on the Digital Twin.

  3. Analyze the results: All parameters of the process and simulation scenarios can be compared as a before/after analysis.

Use Cases

  • S/4HANA Transformation Impact
    • See how it will affect your process KPIs and impact your business
  • System Migration Impact
    • Calculate the impact of automation or new features on your process KPIs
  • What-If Simulations for M&A Synergies
    • Effectively realize synergies in aligning resources and process flow
  • Shared Service Transition
    • See where in the process to best invest into more resources
  • Throughput vs. Cost Optimization
    • Simulate how big or small process changes impact process KPIs
  • RPA/Automation Prioritization
    • See if it really speeds up the process, or just shifts the bottleneck somewhere else
  • Process Capacity Limit Identification
    • Simulate what happens if a huge number of cases enters your process at once

How to use Process Simulation in python

To get started, we have updated the pycelonis documentation with tutorial notebooks on Process Simulation.
Note: You need to update pycelonis before you can use the new functionality.

Currently, we include two tutorial notebooks:

  1. Introduction to the Process Model
  2. Celonis Process Simulation Tutorial (main tutorial)

To import these notebooks into your ML Workbench, run:

from pycelonis.notebooks import process_simluation

The first notebook gives you an introduction to the parameters of a simulation model. Those include information about the resources, processing times, and many more. The process model (or Digital Twin) builds the foundation for all the process changes you want to simulate.

Find it here

The main tutorial covers a full end-to-end demonstration of the simulation capabilities:

  • Extract the as-is Process Model from the data: In this step, we are semi-automatically extracting the Digital Twin from your data model. Some parameters can be extracted fully automatically, others need to be validated by the user or input manually.
  • Simulate the as-is Process Model: After extracting the Digital Twin, we are going to execute a simulation. This gives us the baseline results that we are going to use in a comparison with another simulated scenario
  • Modify the as-is Process Model to try out alternative models: In this step, we are going to make changes to the Digital Twin, and thus create a second simulation scenario
  • Simulate the alternative models: Simulate the second scenario and compare the results to the Digital Twin

Find it here

With that, we wish you happy holidays! We are hoping you find this functionality useful and are looking forward to your feedback.

Best Regards,

Celonis Product Team

2 Likes