Goldratt SIM - Gas Station

macOS / Economie et entreprise

At Goldratt Research Labs we believe small changes can unlock BIG results.

We collaborate with individuals and organizations to discover and release their inherent potential through small changes that are simple, fast and low risk.
Complexity and uncertainty faced by organizations and individuals make it very difficult to decide which small changes will bring the BIGGEST results.
We develop decision support and simulation tools that help our customers make better, faster decisions and predict the impact of their decisions and changes in the environment.
Goldratt Gas Station Simulation is a simple simulation that illustrates how difficult it is to predict the impact of decisions on an interdependent​ system


How will the service type offered to customers impact the financial and operational performance of a simple gas station?

The main objective of this model is to determine how different customer service types impact the average service
time and therefore the customer experience at the gas station.
There are 5 customer service types that we want to model:
- Self-service station (no attendants), payment before filling
- Fully automated gas station with self-payment terminals
- Full-service gas station, attendant fills, payment before filling
- Self-service station (no attendants), payment after filling
- Full-service gas station, attendant fills, payment after filling

We also want to determine where the bottleneck is in the system for each of the customer service types. It can be whether or not we have the capacity at the cash desk or whether we have enough attendants to service customers. There are sliders to change the values of these parameters.

Even a simple system like a gas station that only consists of queues and process delays have non-linear relationships and trying to solve these questions using deterministic spreadsheet-based tools will be inadequate. A more suitable way to model systems is a simulation as it provides a means to model interdependencies between processes and measure in detail the impact of change on the system.

We are able to run the simulation model as an individual experiment and view the animation or we can run multiple
experiments and compare the outputs next to each other.