The following Examples show what Perfomance Indicator can be used for, and show what goals we actually want to achieve. All of them can be implemented using the classes found here.
In some Application there are some parts ordered externally Orders For every order ther is a expected (wanted) price and the price finally made by the external supplier. This is a global Perfomance indicator with to bases: expected and actual.
In some large Project there shall be a monthly. statistic for all parts of the projcts about the number of people involved. So we crear a Perfomance Indikator (PI) "numPeople" for the Poejcts its Subprojects, Workpieces and so on. The PIS will have monthly values. The values can only be set after the month has finished. The values must be agregated as sum over all. The base is simple (but needs an additional inner base for the calculated values). The aggregated value can only have a meaning when all values are set so the final value will only be aggregated when all "lower" values are set.
While planning a Project every part of the Project has an expected Cost. While the Project is running the actual costs are found and entered. Now we can compare expected and actual costs. So the bases are expected, actual (and an internal one for calculated values.). For a Complete financial analysis these costs must be subdivided into Costtypes. Costtypes again are structured into Costgroups. These need a second type of Aggregations to build sums for the Cosstypes and Costgroups (over all Projects) a GUI must allow a drill down to all details.
For most Part of the Project there are status reports to be filled. So for every status report we want a state of "green", "yellow" or "red". To get more detail these states are aggregated from other, typed states like Time, Quality, Budget. The aggregation is customizable so that the calucaltion of the main status ist customizable. In Addition the status reports of Projects higher in the hierarchy are influenced by thier subprojects. Statusreport may not fit into a even time scale (Weekly reports may be dropped during some longer holiday. There may be to sucessing, but different reports for the same month) So aggregation may become difficult.
For some large scale movie the Productmanager wants to plan and monitor the amount of resources. As usually for a movie for every sevente there will be several takes and only a cutot from these will make int intol the final movie. So he plans the following indicators for every scene: planned scene-time, planned take-times, final scene-time, final take-times. The time is measured in seconds. Depended on the type of Scene (Mass/Private, exterior/interior, cgi, ...) This could be used to calcualte a rough cost estimation. Aggregation needs no internal base since only the leaf nodes can have values.
Copyright © 2005 Business Operation Systems GmbH