190
6 Experimental Validation of the Methodology
6.4
Recommendations for Action for Bosch Derived
from the Simulation-based Optimization
Methodology
In the previous sections, it was demonstrated that the simulation-based optimization methodology is suitable for practical application. Basically, the creation
of a simulation model requires the existence of the corresponding data. In addition to the generally needed data for production planning and control, the energy
consumption data of the machines must be available. It has been shown that the
resolution of the underlying data base for the energy consumption profiles should
be chosen in a way that it is accurate enough to represent the consumption behavior and the load peaks of the machine states. In the case of Bosch, it has been
shown that the resolution of the energy data provided on the energy platform is
not suitable for use in the simulation model to conduct optimization studies. Due
to the selected 15-minute interval, the energy states cannot be clearly recognized
and separated from each other in the energy data. Thus, a realistic reproduction
of the energy consumption behavior of machines in the simulation model is not
possible.
It is therefore recommended to use energy data with a resolution of one data
record per minute. Should data protection regulations prevent such a high resolution from being made available (temporarily or permanently), machine state-based
mean values should be calculated on the basis of the 1-min resolution data and
used in the simulation model instead of the exact energy profiles, since good
results can be achieved with correctly calculated mean values, at least with regard
to the total consumption optimization.
The use of different information technology (IT) systems to document different production and machine data has proven to be disadvantageous, as
non-synchronized time stamps and different data acquisition methods require high
manual efforts for the merging of production and energy data. The documentation of all data in one single system having the big advantage of one consistent
time used for all data records, would considerably simplify and accelerate the
connection of relevant information. It is recommendable to integrate the documentation of production quantities, machine failures and maintenance plans in
the ERP/MES system instead of maintaining them in several different applications. The parameter configurations determined as optimal within the framework
of the methodology can only be used as recommendations for action to switch off
machines during production shutdowns if they are based on a correct and reliable
data basis.
6 Experimental Validation of the Methodology
6.4
Recommendations for Action for Bosch Derived
from the Simulation-based Optimization
Methodology
In the previous sections, it was demonstrated that the simulation-based optimization methodology is suitable for practical application. Basically, the creation
of a simulation model requires the existence of the corresponding data. In addition to the generally needed data for production planning and control, the energy
consumption data of the machines must be available. It has been shown that the
resolution of the underlying data base for the energy consumption profiles should
be chosen in a way that it is accurate enough to represent the consumption behavior and the load peaks of the machine states. In the case of Bosch, it has been
shown that the resolution of the energy data provided on the energy platform is
not suitable for use in the simulation model to conduct optimization studies. Due
to the selected 15-minute interval, the energy states cannot be clearly recognized
and separated from each other in the energy data. Thus, a realistic reproduction
of the energy consumption behavior of machines in the simulation model is not
possible.
It is therefore recommended to use energy data with a resolution of one data
record per minute. Should data protection regulations prevent such a high resolution from being made available (temporarily or permanently), machine state-based
mean values should be calculated on the basis of the 1-min resolution data and
used in the simulation model instead of the exact energy profiles, since good
results can be achieved with correctly calculated mean values, at least with regard
to the total consumption optimization.
The use of different information technology (IT) systems to document different production and machine data has proven to be disadvantageous, as
non-synchronized time stamps and different data acquisition methods require high
manual efforts for the merging of production and energy data. The documentation of all data in one single system having the big advantage of one consistent
time used for all data records, would considerably simplify and accelerate the
connection of relevant information. It is recommendable to integrate the documentation of production quantities, machine failures and maintenance plans in
the ERP/MES system instead of maintaining them in several different applications. The parameter configurations determined as optimal within the framework
of the methodology can only be used as recommendations for action to switch off
machines during production shutdowns if they are based on a correct and reliable
data basis.