Integral Logistics Management — Operations Management and Supply Chain Management Within and Across Companies

14.3.1 Operations-Oriented Finite Loading, or Operations Sequencing

Intended learning outcomes: Produce an overview on the technique (the algorithm) for operations-oriented finite loading, also called operations sequencing. Explain an example of operations-oriented finite loading. Describe various priority rules. Present an evaluation of the technique. Identify its limitations and typical areas of application.



Operations-oriented finite loading aims to minimize possible delays to individual operations and thus the average potential delay of the entire production order. 
Operations sequencing and operations-oriented finite loading are synonymous. 

Overview: The individual operations are planned time period by time period on the basis of orders, starting from the start date determined by lead-time scheduling (Section 13.3.3).

Planning strategy: This means establishing meaningful rules of priority for the order in which operations are scheduled, with the aim of achieving maximum throughput. The queues waiting upstream of the work centers are monitored and adjusted.

Technique: The planning horizon is divided into time periods. The operations to be scheduled are then assigned to work centers, period by period, until the capacity limit is reached, regardless of the order to which they belong. Figure 14.3.1.1 demonstrates the principle of the resulting algorithm. This includes the following aspects:

Fig. 14.3.1.1       Technique (algorithm) for operations-oriented finite loading.

  • Work center priority: The order of the work centers becomes im­por­tant as soon as there is more than one operation to be scheduled for an order in each time period. Possibly, the subsequent operation then relates to a work center whose planning has already been carried out for this period and now must be revised.
  • Determine the operations to be scheduled in the first time period; typical operations are, firstly, every (subsequent) operation waiting for execution, for orders already started (the data on order progress identifies these operations), as well as, secondly, every first operation for orders not yet begun whose start date — calculated using a scheduling method (Section 13.3) — lies within the first time period.
  • Determine the operations to be scheduled in time period i, 2£ i£ n; candidates are, firstly, all operations not scheduled in the previous time periods; then, secondly, those operations for which the previous operation was scheduled in an earlier time period and whose start dates lie within time period i, as well as, thirdly, every first operation for orders not yet begun whose start date — calculated using a scheduling method (Section 13.3) — lies within the time period i.
  • Arrange the plannable operations by priority. The following secondary objectives may be applied to the selected order:
    A.   Minimize the number of delayed orders
    B.   Apply an equal delay to all orders
    C.   Minimize the average wait time for operations
    D.   Minimize the number of orders in process
  • The following priority rules may be applied (see also [RuTa85]):
    1.    The order in which the operations arrive (FIFO, “first in, first out”)
    2.    Shortest processing time rule (SPT)
    3.    Proximity of the order due date (EDD, earliest due date)
    4.    The ratio “remaining lead time for the order divided by the number of remaining operations”
    5.    The ratio “remaining lead time for the order divided by the time still available for the order” (SLK, shortest slack time rule, » order urgency; see also Section 13.3.6)
    6.    The ratio “remaining lead time for the order divided by the remaining operation time for the order”
    7.    (External) order priority
    8.    Any combination of the above

    Rules 1 and 2 are the easiest to apply in control of operations, be­cause the informa­tion is immediately available: it is physically visible “locally.” It is not necessary to consult a computer or a list. The other rules may require complicated calculations.

    Every priority rule takes into account one or another secondary objective. Rule 1 is often used, since it minimizes the wait time upstream of the work center and thus the average order delay (objectives A and B). If capacity is utilized more fully, the strategy changes, and rule 2 is chosen. This accelerates the largest possible number of orders and thus reduces the value of goods in process (objectives C and D).
  • Load the operations in order until the capacity limit is reached: If an operation exceeds the capacity limit, we transfer any as yet unscheduled operations to the next time period. The capacity used for the overlap load for the last operation is then no longer available in the next time period.

    One variation is not to schedule the operation that exceeds the capacity limit. However, this will use up remaining capacity only if an operation with a smaller load can be scheduled. This variation requires a more complicated algorithm.
  • Calculate the start date for the next operation: After loading the operation, we calculate its completion date and the start date of the next operation on the basis of the inter­operation time. To avoid problems with the algorithm (see “priority of the work centers” above), it may be useful to use the start of the next time period as the earliest start date.[note 1407]

Figure 14.3.1.2 shows the result of operations-oriented finite loading using the orders in Figure 14.2.1.1, specifically P1, . . . , P6, and the same work centers, namely, work center A and work center B. Priorities were assigned in ascending order of order ID. Again, “preload” represents operations for orders that were loaded before orders P1, . . . , P6.

Fig. 14.3.1.2       Example of operations-oriented finite loading.

In contrast to the load profile in Figure 14.2.1.1, in finite loading we display the loads rotated 90° toward the time axis, whereby the height of the bar is equal for all work centers. The period length is then standardized at 100% capacity over the time period. This technique is possible because the load does not usually exceed capacity. We can then enter a number of work centers along the vertical axis. Utilization of the entire system is evident at a glance.

Evaluation of the technique: The following prerequisites must be met to use this technique:

  • Capacities and loads must be sufficiently reliable, that is, the planning data and reported work progress must “tally.” Other­wise, errors can accumulate very rapidly in the calculated dates.
  • Due dates must be sufficiently flexible: We set the completion date for an order randomly on the basis of the existing utilization of production capacity. Lead times can be considerably longer than originally planned, however.
  • It must be possible to limit the optimization of set-up times to the operations within a given period.

This creates the following limitations:

  • The further we plan into the future, the smaller our chances that the planning forecasts will prove correct, if only due to unforeseen breakdowns or incorrect load specifications. For this reason, the technique is only sufficiently exact for short planning horizons, and it must be repeated at regular intervals.
  • To be able to work to schedule in subsequent periods, any scheduled operations must be completed during this period. The technique does not allow reactive replanning locally.
  • The level of goods in process is of secondary importance, both financially and with respect to volume. The planner monitors and adjusts the queues upstream of the work centers. Capacity is relatively inflexible, however, so orders must be held back, i.e., not released in good time. With long lead times in particular, however, order release can occur at the first identification of a bottleneck. This will physically hold up the production plant. Choosing a “neutral” priority rule will distribute the delay more or less evenly among all the orders.

The following are the typical areas of application:

  • For batch productionover a long period or in a monopoly situation; that is, in a seller’s market. In such cases the date of delivery, for example, to the end products store or to the customer, is less important. Some typical industries that belong here today are the chemical and food processing industries and niche capital goods markets.
  • The operations-oriented finite loading technique simulates a situation that may arise in job shop or even line production. The operations for an order are executed in a more or less random order, in competition with other such orders. For execution and control of operations, this type of planning provides a process simulation for the coming days and weeks; that is, an actual working program for the shop floor.



Course section 14.3: Subsections and their intended learning outcomes

  • 14.3 Finite Loading

    Intended learning outcomes: Explain operations-oriented, order-oriented, and constraint-oriented finite loading.

  • 14.3.1 Operations-Oriented Finite Loading, or Operations Sequencing

    Intended learning outcomes: Produce an overview on the technique (the algorithm) for operations-oriented finite loading, also called operations sequencing. Explain an example of operations-oriented finite loading. Describe various priority rules. Present an evaluation of the technique. Identify its limitations and typical areas of application.

  • 14.3.2 Order-Oriented Finite Loading

    Intended learning outcomes: Produce an overview on the technique (the algorithm) for order-oriented finite loading. Describe various priority rules as well as dealing with exceptions. Explain an example of order-oriented finite loading. Present an evaluation of the technique. Identify its limitations and typical areas of application.

  • 14.3.3 Constraint-Oriented Finite Loading

    Intended learning outcomes: Identify bottleneck capacities and the drum-buffer-rope technique. Describe the drum, the buffer, and the rope. Present an evaluation of the technique. Identify its limitations and typical areas of application.