The planning algorithm tries to calculate the perfect cleaning schedule for the next execution every 30 minutes. The algorithm takes different factors into account. All services of surfaces that are not scheduled are kept in the Backlog.

The planning algorithm distinguishes between restrictions (hard specifications), which the algorithm must fulfill, and weightings, which help the algorithm to find the perfect solution.

Planning algorithm

Restrictions

Restrictions	Description
Maximum cleaning tour duration	A cleaning tour has a maximum duration, which is set by the user. This duration will not be exceeded.
Minimum number of area treads	The number of cleaning tours entering an area is reduced to the minimum.
Services on the surface are not separated	The services of a surface, which can be executed by a cleaning tour, are completely scheduled or not scheduled. A separation does not take place.
Columns cleaning	If the column cleaning is activated on categories, all services from these tours will be completely scheduled or not scheduled.
Category may not be planned	With the restriction Services on the surface are not separated, all services of a surface are always planned or not planned. With this option, which can be set on the category, a category can be excluded from this restriction. For example, this can be used for monthly cleaning with its own service number.
Separate maximum duration for substitution	If the maximum time for the deputy tasks is to be handled separately, the maximum duration for these tasks is handled in this restriction.
Surface group can only be assigned on a cleaning tour	The GeoClustering algorithm creates area groups that are ideally processed by one tour. This restriction prevents several cleaning tours from cleaning in the same surface group.

Weightings

Weighting	Description
Priority	Priority of the services. The higher the priority, the higher the weighting.
Day Priority	For services that can only be performed today, an additional priority is charged. Such services are daily fixed services or services which have the max. weekly frequency.
Minimize cleaning tours on floor	As few cleaning tours as possible should enter a floor.
Minimize cleaning tours in buildings	As few cleaning tours as possible should enter a building.
Minimize floor changes and floor jumps	A tour should make as few floor changes as possible and the floor changes should be as small as possible.
Homebase	If a homebase is set for a cleaning tour, the cleaning tour should clean as many surfaces as possible that are in the vicinity.

Surface groups (GeoClustering)

If several cleaning tours can perform the same surfaces on one floor, these surfaces are grouped together and a group is assigned to only one tour. The so-called GeoClustering has the effect that individual, distributed surfaces are not scheduled on a floor during cleaning tours.

This function is only active if the business entity has been set up with building plans.

Example: Two tours share a floor and can clean all surfaces on this floor. 4 surfaces are due for cleaning. Three surfaces are on the left side of the floor, one surface is on the right side of the floor. The GeoCluster algorithm will form face groups based on the geometry and number of tours. In this example, the three faces into one face group, one face into the other face group. One face group is assigned to the first tour, the other face group is assigned to the other tour, or both face groups are assigned to one tour.

Priority

A basic priority can be defined on the area type. This is the priority that is used as soon as a service is due on an area. This priority increases continuously as long as the area is not cleaned. This prevents surfaces with a low priority from never being cleaned when the tour load is high. If a surface is overridden by the user with now clean, this surface gets the highest possible priority.