Dynamic Discrete Event Modelling for Operations

We are currently working on developing a dynamic discrete event model for operations processes. This is intended to move us one notch closer towards having the ability to create a digital twin for everything operational.

For this kind of modelling the corresponding trigger events (be it an incident or a fulfillment request) are individually created with type and frequency described by statistical distributions. Each of the Events is then tracked through the modelled process on its way occupying limited resources that can be limited (which requires queuing ahead of the actual utilisation of the resource) and can be related to aspects like cost, energy consumption, pollution etc.?

Having such a model available enables much more detailed diagnostic data to be extracted from (simulated) process execution. To achieve a similar level of detail in real life process execution would require massively instrumenting the process which (at least for partially manual process implementation) leads to additional effort and is thus significantly impacting process execution thus creating systematic measurement errors.

The insights derived for the diagnostics are helping to identify the most relevant candidate domains for improvement in an unprecedented level of detail enabling far better targeting of improvement initiatives combined with modelling-based quantification of the expected improvements in metrics.

We intend to utilize this model to gain insights into the effects of levels of autonomy, green operations and several other current operations topics.

A more visionary option of use for this type of modelling initiative is the creation of a digital twin of an actual operations organisation (using actual events to trigger the processes), enabling the identification of deviation of actual performance from expectation, triggering corrective action, in addition to enabling more insights without adding administrative burden.

Happy to read from you!