7.2 Critical Appraisal of the Methodology
199
In section 6.5 the simulation-based methodology is evaluated with regards
to the practical applicability. While several requirements are included and fully
met by the simulation-based optimization methodology, for example, the transparent presentation of interactions of single production aspects, the modular
structures with a high number of parametrization options and the included optimization experiments that can be executed without requiring complex interface
management to link different software solutions, some concept objectives and
requirements remain unsolved. On the one hand, the commitment to AnyLogic
meets the requirement of having one single software solution supporting both,
multi-method simulation and optimization experiments. On the other hand, limitations regarding the number of supported database sizes, the number of possible
optimization parameters, and allowed iterations in the optimization experiments
arise. The coupling of a multi-method simulation software with an optimization
software could eliminate those limitations but results in higher efforts for interface management and data-exchange and thus, probably demands for a simulation
expert.
The simulation-based optimization methodology proposes optimized parameter
configurations for energy-efficient machine control. The idle and standby optimizer parameters define from what production interruption lengths on it is more
efficient to turn the machine off rather than having them run in idle or standby
state. However, to bring these parameter configurations in the real production, the
employees need to know at the beginning of each interruption how long it will be.
Depending on the causes for the production break down, the interruption length
cannot be estimated in advance. While the point in time when a restocking of the
buffers will end a material shortage might be known from supplier information
or planned material deliveries in the ERP system, machine break downs add an
unknown factor to the system. If production interruption durations can only be
roughly estimated, the worker has to decide on his own, if he follows the proposed parameter configurations and switches the machines off or if he leaves them
in idle mode in case the interruption might be too short to invest the energy and
time required for the warmup of the production lines. It can be summarized, that
the total consumption optimization requires the knowledge of interruption lengths
to make use of the optimization results.
The optimal value for the offset parameter on the other hand is determined
within a range of a few seconds to a few minutes to avoid negative effects on
the production output. The applicability to manually switch machines on and off
that precise highly depends on the level of automation in production. To realize
the recommendations for the machine offset, it requires an automated mechanism,
199
In section 6.5 the simulation-based methodology is evaluated with regards
to the practical applicability. While several requirements are included and fully
met by the simulation-based optimization methodology, for example, the transparent presentation of interactions of single production aspects, the modular
structures with a high number of parametrization options and the included optimization experiments that can be executed without requiring complex interface
management to link different software solutions, some concept objectives and
requirements remain unsolved. On the one hand, the commitment to AnyLogic
meets the requirement of having one single software solution supporting both,
multi-method simulation and optimization experiments. On the other hand, limitations regarding the number of supported database sizes, the number of possible
optimization parameters, and allowed iterations in the optimization experiments
arise. The coupling of a multi-method simulation software with an optimization
software could eliminate those limitations but results in higher efforts for interface management and data-exchange and thus, probably demands for a simulation
expert.
The simulation-based optimization methodology proposes optimized parameter
configurations for energy-efficient machine control. The idle and standby optimizer parameters define from what production interruption lengths on it is more
efficient to turn the machine off rather than having them run in idle or standby
state. However, to bring these parameter configurations in the real production, the
employees need to know at the beginning of each interruption how long it will be.
Depending on the causes for the production break down, the interruption length
cannot be estimated in advance. While the point in time when a restocking of the
buffers will end a material shortage might be known from supplier information
or planned material deliveries in the ERP system, machine break downs add an
unknown factor to the system. If production interruption durations can only be
roughly estimated, the worker has to decide on his own, if he follows the proposed parameter configurations and switches the machines off or if he leaves them
in idle mode in case the interruption might be too short to invest the energy and
time required for the warmup of the production lines. It can be summarized, that
the total consumption optimization requires the knowledge of interruption lengths
to make use of the optimization results.
The optimal value for the offset parameter on the other hand is determined
within a range of a few seconds to a few minutes to avoid negative effects on
the production output. The applicability to manually switch machines on and off
that precise highly depends on the level of automation in production. To realize
the recommendations for the machine offset, it requires an automated mechanism,