Occasionally in the real world delivery vehicles get delayed and don’t arrive on the day they were scheduled to arrive. In the SCM Globe simulations it happens that way too. As you get familiar with the simulation and the data that is created, you will notice that sometimes a scheduled delivery doesn’t happen. That’s not a bug, it’s another feature to add realism to the simulation… seriously… 🙂
It is not realistic to operate a supply chain as if everything will happen every day just as planned. That’s why supply chain managers in the real world keep safety stock inventory at facilities. Safety stock enables the facilities to operate even if needed supplies do not arrive as planned. Usually safety stock at a facility is set to the amount of inventory used during one delivery cycle. If deliveries are made every day, then safety stock is one day’s worth of inventory usage. If deliveries are made every three days, then safety stock is three days worth of usage, and so forth. But remember, the goal is still to minimize on-hand inventory and safety stock inventory so don’t let too much safety stock build up at any facility in the supply chain.
In the simulation logic for a vehicle there is a variable for the delay between departures. If you set that delay to 24 hrs, that means that the vehicle will run its route and return to its start location and depart again 24 hours later. If the route takes 4 hours then the simulation begins calculating the next departure as 24 hours after the completion of the 4 hour delivery route. So the departure of the vehicle keeps getting pushed up by 4 hours every time the vehicle runs its route. In this case, after six departures and the addition of 6 x 4hrs = 24hrs to the time between departures, it pushes the departure into the following day. So a departure does not happen at all for that vehicle that day.
This means that the longer it takes a vehicle to complete its route, and the more frequently it travels the route, the more frequently there will be a missed delivery. And that same logic will also, but less frequently, cause deliveries to sometimes be made twice when only one was expected. So a bit of extra storage space at facilities is needed to handle this. The baseline missed delivery rate modeled in SCM Globe may be somewhat higher than what occurs in most real world supply chains, yet this does not change the validity of the simulation results. A supply chain design that performs well even with high missed delivery rates is indeed a good design. When you find a good design, you can further refine it by reducing missed delivery rates.
If you want to reduce the missed delivery rate on a route, subtract the travel time for the vehicle on that route from the delay between departures for the vehicle. When you edit a route, you can see the time it takes to travel the route. In this case where the vehicle takes 4 hours to travel its route you can subtract 4 hours from the delay between departures and get a daily departure on all days of the simulation: 24 – 4 = 20; set delay between departure to be 20 hrs.
There are inevitable rounding errors since the simulation algorithm calculates out to 14 decimal places and we only display 2 or 3 decimal places onscreen. In the thousands of calculations that occur during simulations, small rounding errors can accumulate and change your expected results by an hour or two one way or the other (this is the Butterfly Effect at work). If your first adjustment doesn’t eliminate missed deliveries over a 30 day period, then experiment with different values for the delay variable. Using the example above, if 20 hrs doesn’t quite work, then try 19 hours, or 19.5 hours, or 21, etc.
Think more about reducing percentages of missed deliveries than about trying to eliminate all missed deliveries. It is more realistic to create supply chain models with low percentage rates of missed deliveries than to create models with absolutely no missed deliveries. Such perfect supply chains do not exist in the real world. The screenshots below illustrate the effects of percentage reductions in missed deliveries.
The graph on the left above shows the baseline missed delivery rate in a simulation using a 24 hour delay between departures on a particular route. On the right above the graph shows the result of a 50% reduction in missed deliveries (subtract half the travel time on the route from the delay between departure). Notice how on-hand inventory levels fluctuate inside a lower range: 25 to 55; versus 30 to 60. And see how average on-hand inventory over the entire period is also lower in the graph on the right. Due to fewer missed deliveries, delivery quantities can be smaller making average on-hand inventory lower. Delivery quantities can be smaller because fewer missed deliveries means there is less need to continually replenish safety stock.
Above on the left is the result of a 75% reduction in missed deliveries (subtract three quarters of the route travel time from the delay between departures). This causes on-hand inventory to fluctuate within an even smaller range, and the average on-hand amount over the entire period continues to drop. On the right is a 100% reduction in missed deliveries with no fluctuation in on-hand amounts over the entire period. Notice how on-hand inventory level is maintained at 30 because it is the appropriate safety stock level when deliveries are made on a daily basis and daily product demand equals 30.
This sequence of on-hand inventory graphs illustrates why as the number of missed deliveries decreases, it is possible to lower on-hand inventory. And with less on-hand inventory it is possible to also reduce related rent and operating costs associated with that inventory.
Copyright © 2016 by SCM Globe Corp.