Littlelabs Simulation

CAPACITY UTILIZATION:

Our team’s overall strategy for managing capacity was to monitor the queue and utilization of each station. We had a bit of a late start, which combined with an adjustment period to the use of the simulation, quickly led to a large queue for the first station. At its peak, there were 300 units backlogged. Given the cost of machines, we took a perhaps overly conservative approach to adding machines, at least in the beginning of the simulation. While we were at 100% utilization for over 100 days early in the simulation, we recognized that we would be at 100% utilization until we cleared out the queue and did not want to overcorrect by creating excess capacity. After stabilizing station 1, it was relatively simple to monitor the utilization of each station and add machines when utilization was consistently over 80%. The key here is consistently, as the utilization naturally varied across a range, so adding machines after a single instance of exceeding 80% utilization could have been unnecessary. We also monitored trends with incoming orders, and knowing there was an expectation that demand would dramatically drop sometime after day 300. We understood that the sale value of the machines was the same regardless of the time of sale, so we held most of the machines as long as possible, but quickly sold as many as possible as utilization dropped and we approached the end of the simulation.

INVENTORY:

With an ordering cost of $1000, at first the team strategy was to minimize reorders. We forecasted our demand for the 14-day lead time for the order to come in, using an estimate of the 50 units average demand quantity per day, multiplied by 14 days, for a total of 700 units. Order quantity was set to 750, and the re-order quantity was set to 700. With a couple of near misses where our inventory when down to 7 and then 2, we re-evaluated our strategy, and added a buffer, raising our order quantity to 1200 and then 1500 units, and our re-order point to 800, in anticipation of the peak demand starting around day 240.

We monitored the demand to determine

...