Opening a warehouse is a long and challenging process, not to mention costly. So if you're going to go through all that work to open and operate a warehouse, you'll want to make sure you can run it as efficiently as possible. Enter SimPy (rhymes with "Blimpie"). SimPy is a discrete event simulator software library that allows us to asynchronously schedule and process events.
This talk will introduce SimPy and how it can be used to build simulations of complex logistics machines using Rent the Runway and it's recent second warehouse opening as a case study.
SimPy is an object-oriented process-based discrete-event simulation framework based on standard Python. At its core, it allows us to asynchronously schedule and process events. We will delve into how SimPy processes yield events and consume resources and also how we monitor them.
Then we'll discuss how we put the simulator to work. Some people think we build simulations to predict the future. However, this is rarely the case. We create simulations to understand a process better, learn more about a hypothesis, or test an intervention. At Rent the Runway, we used SimPy to help us understand how our strategies would affect the opening of our second warehouse like which customers would be served by which warehouse and how would we divide inventory between the two.