| Imagine that you're the manager of a department | | | | process from when the mill ordered a train to |
| in a railway that rails sugar from the mill to the | | | | when the train left the port after unloading the |
| port. Your department takes orders from the mill | | | | sugar for export. An important output for the |
| for trains when they're needed, to come and load | | | | sugar train system is the number of tonnes of |
| up with sugar at the silos, and chug them along | | | | sugar delivered to the port on time. In a way, it's |
| the tracks to dump the sugar at the port for | | | | a measure of the capacity or throughput of the |
| export. | | | | train system. |
| | | | | |
| And you've just had a call from the manager at | | | | #3: In-Process Measures - these are the powerful |
| the sugar mill: "Where are those *%$# trains we | | | | measures! |
| ordered?! We've had to shut down production | | | | |
| AGAIN because the silos are full. You know how | | | | One of the problems in the sugar train system |
| much that costs us! When are you guys going to | | | | was that the trains, in fact, were not keeping up |
| get your train problem sorted?!" | | | | with production. And the railway's first response - |
| | | | | by tradition - was to put more trains into the |
| It's not too far from the truth. Many years ago I | | | | system. Very expensive. But our forward-thinking |
| worked with a pretty forward-thinking manager in | | | | sugar train manager looked deeper into the |
| the railways, and one of his processes was the | | | | problem. |
| flow of sugar from mill to port. And he probably | | | | |
| had not too dissimilar words with the sugar mill | | | | He discovered that increasing the number of |
| manager on the odd occassion. | | | | trains or the number of wagons on each train |
| | | | | would improve capacity at too high a cost. By |
| The problem was in the white space. The hand | | | | modeling the sugar system, he found that he |
| off points between parties in the sugar process, | | | | could increase capacity by using the same (or |
| as sugar flows from the mill to the port. | | | | less) rollingstock in a way that was very different |
| Actually, even as it flows from the canefields to | | | | to traditional thinking: create trains that were a |
| the ships. | | | | fixed length that cycled more frequently through |
| | | | | the sugar system. No trains needed to be |
| And how this railway manager solved the problem | | | | ordered, no need to change the number of |
| was basically looking at the whole process and | | | | wagons on the train each time. The unit-length |
| using three types of measures to find the | | | | trains just kept cycling through the system. |
| bottleneck and fix it. | | | | |
| | | | | And to add insult to injury to traditional railway |
| #1: Process Outcome Measures - these set the | | | | thinking, they made sure they had enough |
| priorities for the process. | | | | wagons on these fixed length trains so that one |
| | | | | wagon would stay empty because it wasn't |
| A really important outcome of the sugar process | | | | needed. Sacrilege! Fancy running an empty wagon |
| was that it flowed without causing the sugar mill | | | | - that's not earning the railway any revenue! |
| to have to stop production on account of not | | | | |
| enough trains to keep emptying the mill's silos. | | | | But that's the key to the transformational |
| Stopping production is a HUGE waste of time and | | | | measure: when that empty wagon was actually |
| capital and labour costs. So this was a very | | | | needed to empty the sugar silos, it was a lead |
| important outcome measure for the sugar train | | | | indicator that production was beginning to exceed |
| process - the hours of mill downtime caused by | | | | the railing capacity. So the railway could ramp up |
| insufficient train capacity. | | | | capacity before the mill ever had to stop |
| | | | | production. |
| #2: Process Output Measures - these help in | | | | |
| diagnosing and testing improvements to the | | | | Transformational measures are almost always the |
| process. | | | | in-process measures. But you have to really |
| | | | | understand your process to find them! |
| The sugar train manager flowcharted the entire | | | | |