Performance analysis
1. Performance analysis
A final aspect to look at is performance, or time in general.2. Performance analysis
Again, we can look into this using some predefined visual techniques, like process maps and dotted charts, as well as through calculating some process metrics. Let's start with the visuals.3. Performance process map
The performance process map is a special type of process map which does not show frequencies on the arcs and nodes, but durations, both of activities and of the times between activities. The type is specified with the performance function, which can be further configured, for example by setting the aggregation that needs to be performed on the durations: mean, median, maximum, etc, as well as the preferred time unit.4. Dotted chart
Another specific technique related to time is the Dotted chart. While the performance process map focusses on the duration of activities, the dotted chart shows the distribution of activities over time. The dotted chart is essentially a scatterplot of activity instances each occurring at a specific time (x-axis) and belong to a specific case (y-axis). By changing the representation of time and the ordering of cases, specific patterns can be observed.5. Dotted chart
This is a dotted chart of the sepsis process we have already seen in the exercises. Here, the x-axis represent the time of the day, while the cases are sorted based on the time of day of their first activity. The dense, sloped line of activities emerges because of the sorting, and shows that starting from 8am in the morning, there are suddenly much more cases appearing compared to the night time. The rate of arrival only slows down slightly towards the evening. Furthermore we see a set of vertical lines consisting of the same activity type, which is represented by the color. This means that these activities always occur at regular intervals: 7am, 8am, 9am, etc. As the day evolves, it can be see that these lines fade away, which means that this pattern is less strong in the afternoon, although not completely vanished.6. Dotted chart
We can look at the same data, but put the time since the start of the case on the x-axis, and sort the cases by duration: we now see the shortest cases on top, and the longest cases below. The view gives an idea about the distribution of the duration. Furthermore, there is a high concentration of the blue activity types within the process, while the pink activities are always found at the end of the process.7. Performance metrics
A more numeric analysis of performance can be done using three different metrics, each expressing a different type of performance: - throughput time is the time since the start of a case until the end of the case, which include both active time and idle time - processing time is the sum of the activity durations, which means it does not include the time in between different activities - idle time is the sum of the durations between the activities, in which no processing of the case takes place. Throughput time tells us how long it takes overall to get something done: for example, how long between ordering a product and delivery. Processing time tells us the real time investment, for example, how much time investment did a particular order ask from all our employees. Idle time focuses on the time a specific case was waiting for something to happen. Each of these metrics works in the way already familiar to us now: using different levels of granularity and the plot function for visualization.8. Let's practice!
Now, let's do some exercises on the time perspective of processes.Create Your Free Account
or
By continuing, you accept our Terms of Use, our Privacy Policy and that your data is stored in the USA.