Parallel Operation

A Parallel Operation is an easy way of modeling several parallel Operations. In fact, internally the the operations of the Parallel Operation object work identically to the standard Operation object, i.e. with individual processing of parts, break downs etc. It is possible to setup a Parallel Operation to process parts in a synchronized manner, in cycles, an example of how this is done is given here.

Note: Settings of the parallel operations are shared, i.e. it is not possible to set individual distributions (e.g. for Process Time or Setup Time) or disturbances for the individual operations.

In the example to the right the two production flows represent the same system as long as the settings of ParallelOperation1 and Operation1-3 are identical and the flow to ParallelOperation1 and Operation1-3 are the same.

Note: The added ease with which parallel operations can be modelled like this comes at the cost of not being able to direct variants to the individual operations within the Parallel Operation.

States of the Parallel Operation


Flow Objects can be in different states during a simulation run and at the end of the simulation this is presented as the utilization of the object.

Intermediate state

For the Parallel Operation there is no direct way of aggregating the states of the individual operations into a state of the Parallel Operation. For example, what should be the state of a Parallel Operation with 2 Operations, one Working and the other one Failed?

In FACTS the following aggregation/priority is used (note the order is important):

  1. Unplanned & Paused: Common for all individual operations, i.e. also for the Parallel Operation.
  2. Failed: If all individual operations are Failed the Parallel Operation will be in the Failed state.
  3. Waiting: If at least one of the individual operations are waiting the Parallel Operation will be in the Waiting state.
  4. Working: If at least one of the individual operations are Working the Parallel Operation will be in the Working state.
  5. Setup: If at least one of the individual operations are Setting up the Parallel Operation will be in the Setup state.
  6. Blocked: If at least one of the individual operations are Blocked the Parallel Operation will be in the Blocked state.

The intermediate state is only used for Animation purposes.

Summarized states (Utilization and Bottleneck data)

The intermediate state is not used for the Utilization-chart. Instead the time spent in the different states by the individual operations are weighed together to form the utilization of the Parallel Operation. The same goes for the bottleneck data (sole and shifting portions), i.e. for the Parallel Operation these are averages of the ones of the internal operations.





Settings

The settings of the Parallel Operation are listed below along with a short description and possibly a link to a more detailed description about the setting.

Setting

Description

Name

Sets the name of the object.

Location

Displays the location of the object.

NumOPs

The number of individual operations within the Parallel Operation. This number is dispalyed over the Parallel Operation as can be seen in the example above.

Note: This can be set to 0, which means that there is no operations in the Parallel Operation. Can be useful for instance in an optimization for determining if an operation is needed at all.

Mixed processing

Determines if parts of different variant types are allowed simultaneously (True) or not (False) on the Parallel Operation.

Note: If False all parts of a certain variant need to leave the Parallel Operation before a part of another variant type is allowed to enter.

Synchronize entries

When True the processing of parts on the Parallel Operation is divided into cycles:

  1. First part enters, and starts an entry timeout (see setting Entry timeout).
  2. During the entry timeout additional parts are allowed to enter.
  3. After the entry timeout or as soon as the Parallel Operation is full a new cycle is started, i.e. the processing of the entered parts are started.
  4. After all parts are finished and have left the Parallel Operation a new cycle can start, i.e. start a 1. again.

Entry timeout

The time during which parts are allowed to enter the Parallel Operation before a new cycle is started, see setting Synchronize entries. Used to increase the likelyhood of having a full cycle, i.e. a cycle where all internal operations are occupied.

Synchronize exits

If True all occupied internal operations of the Parallel Operation need to be finished and operational before any part is allowed to leave the ParallelOperation.

Note: This will only ensure that all parts will try to leave together, if one or several are blocked they will not leave at the same time.

Failure zone

Used to link failures between the internal operations of the Parallel Operation, i.e. if True and one internal operation breaks down all others will too.

Note: When used this will result in a lower availability than the one specified for the Parallel Operation.

Process time

The process time of the Parallel Operation models the time it will take the individual Operations to process a variant/product. See process time for more details.

Process time EPT

Used to add additional variations to processing time. These variations could be due to natural variations in execution time of the task and be set to {none, manual}. Default this is set to none.

None: No additonal variation is added.

Manual: This adds variations that will mimic variations seen on an task performed by a human. These variations include all types of disturbances, both failures and human variations in execution of the task. Variations in human execution might for example be due to talking to a coworker or going to the toilet.

This is done using something called Effective Process Time (EPT), which in this implementation is based on real data from a case performed in the automotive industry.

Note: Disturbances set this way will be added to the processing time and hence be logged as working in the output, see the output utilization.

Setup time

The setup time of the object models (if used) times for setting up the object for an other variant than the previously processed one. See setup time for more details.

Note: Each individual operation will have their own setup, i.e. they can be setup for different variants.

Setup time EPT

Used to add additional variations to setup time. These variations could be due to natural variations in execution time of the task and be set to {none, manual}. Default this is set to none.

None: No additonal variation is added.

Manual: This adds variations that will mimic variations seen on an task performed by a human. These variations include all types of disturbances, both failures and human variations in execution of the task. Variations in human execution might for example be due to talking to a coworker or going to the toilet.

This is done using something called Effective Process Time (EPT), which in this implementation is based on real data from a case performed in the automotive industry.

Note: Disturbances set this way will be added to the setup time and hence be logged as setup in the output, see the output utilization.

Disturbance

A detailed description about disturbances can be found here.

Note: Each individual operation will fail independently of the other operations within the Parallel Operation.

Exit Strategy

The exit strategy sets how variants/products should exit from the object to its successors. See exit strategy for more details.

Dispatch

The dispatch setting sets how variants/products waiting to enter the object should be prioritized. See dispatch for more details.

Timetable

If the object follows a certain Timetable it will be shown here. The input-field has a drop-down containing all timetables in the model as well as an option none. Using this drop-down it is possible to select a different timetable or remove the timetable from the object by selecting none.

User attributes

Lists all attributes that the user has added to the object. See User attributes for more details.