Visual Planning and Scheduling Systems A White Paper By Gregory Quinn President Quinn & Associates Inc September 2006 Updated October 2006
Visual Planning and Visual Scheduling Systems Introduction The purpose of this white paper is to provide the prospective buyer of a visual planning or visual scheduling system with the understanding of the differences between the two, and why the visual scheduling system provides the superior return on investment of the IT expenditure. Advances in software and small computers have created a powerful combination to create planning and scheduling systems that showcase much improved operator interfaces over their green screen ancestors. This increased performance has allowed systems to become more responsive to real-time problems in the manufacturing and service industries. The new planning and scheduling software systems have begun to blur the distinction between planning and scheduling in the minds of the end users, with good reason, as the graphical user interface for the applications look more similar with each year. Unfortunately, this superficial similarity masks the fact that planning systems have fundamentally remained unchanged from the days of the mainframes, whereas scheduling systems have evolved tremendously to the extent that the visual scheduling system has surpassed the visual planning system in both effectiveness and practicality. Planning vice Scheduling As with any discussion, the definition of terms is essential to form a better appreciation of the issues with planning and scheduling. Enterprise systems have grown to be very complex, and, in the well meaning efforts to make complex systems understandable (and regrettably also due to sales efforts to fit a square peg in a round hole), over simplification has let to some users to think planning and scheduling are synonymous. To review: Planning: To take demand from various sources and to organize a timeline that identifies what items need to be produced, the quantity of each item, and when the items are to be produced over the planning horizon. Scheduling: Take the goals established in the planning cycle and based on the production process and the production environment, determine how demand will be met by establishing the sequence of jobs on each resource over the scheduling horizon. In the typical architecture of an enterprise resource planning (ERP) system, demand is formulated by combining forecast and actual demand and a master production schedule (MPS) is established. While the use of the 1950 s legacy term MPS is literally true, the confusion between planning and scheduling begins here. While the MPS establishes a set of due dates, the data are only demand points in the MPS, and no actual sequence considering the routing steps to produce the end items. Fundamentally the MPS is only a plan.
The MPS then is submitted to the manufacturing requirement planning (MRP) function, which develops a material plan and creates work orders through a bill of material explosion, and this planning cycle may utilize a rough cut capacity plan (RCCP) function. The RCCP takes the manufacturing capacity expressed in hours over some interval (or time bucket) like a week, day or shift; and looking at the time required to manufacture an item determines in which bucket manufacturing for the item will occur. In some cases an algorithm will be used to minimize lateness of the work orders when production exceeds capacity. Generally capacity plans are developed backwards from due date. Once the material and capacity plans are established, then the production schedule can be generated. The key data required for developing the schedule is the routing (in some cases the bill of manufacture which is a combined form of the BoM and routing), which describes each operation step, run times, setup times, and constraint usage. Constraints can be any additional resource such as labor and/or tooling that is required in order to execute a production step. Job release dates can be incorporated into the production schedule, often reflecting time constraints in the system such as material availability. The crucial difference between the scheduling and planning activities is that scheduling is developed in a bucketless timeline to derive precise start and end time for each operation in each work order. A closed loop ERP will feed back the start and end times of each work order step into the MRP to fine tune the material plan to minimize purchased material inventories. By following this flow the reader can see how the overall demand is reduced from general date-based production goals (MPS) to work to be performed during planning intervals (planning) to specific tasks to be performed in defined points in time (scheduling). This process highlights the crucial distinctions between planning and scheduling. Benefits of Planning and Scheduling Before real-time scheduling, capacity planning systems were the most effective means to fit the production requirements within the capacity of the manufacturing environment. Capacity planning is easy, often performed by pencil and paper or using spreadsheets. Capacity planning is better than the infinite loading of resources common to MRP functions. However, the drawbacks of capacity planning are many, rooted in the fact that many generalizations in resources and processes combine to provide at best a first-order estimate if production can be accomplished within the capacity envelope. Further inhibiting the accuracy of the planning process is the idea of time buckets, which bars any optimization of the works orders to best achieve the production goals of the company. Take a simple example of where there are a limited number of technicians available to set up production resources. In a planning system, the lack of visibility within a time bucket will assume that all equipment can be set up to run during that time bucket. In a
scheduling system over the same period of time, only the number of machines equal to the number of setup technicians can be set up simultaneously, reflecting the reality that some operations will be delayed because setups cannot occur. The planning system will indicate no problems with product delivery, an outcome that can be false. A good scheduling system not only rectifies the potential errors from the generalizations of capacity planning, but a real-time scheduler replaces the planning cycle, eliminating a level of data process and creating a more responsive ERP or manufacturing execution system (MES). Quality scheduling systems can simultaneously consider sequencing operations on resources with constraints, as well as going beyond backwards planning, incorporating more complex scheduling objectives such as parallel loading, minimizing work-in-progress (WIP), and minimizing setups and changeovers using sequence dependent setup relationships between products. Furthermore, a scheduling system will report a more accurate cost-of-execution than a capacity plan. The prevailing distinction of a scheduling system when compared to a planning system is the dispatch, or work-to, list. This list describes the sequence of jobs to be run at a work center, denoting not only order of jobs but also projected start and end times. In a bucketed system such as a planning system, it is impossible to determine the order of jobs, thus making the task of maximizing production effectiveness against a stated production goal impossible because there is no way of accurately seeing the effect of problems in a work center and the consequences to the rest of the jobs on the production floor. Jobs are simply in a virtual pile within a time bucket in a planning system. The precision that a scheduling system brings to the production environment provides benefits that are not possible with a time-bucketed system, for example: Increased confidence in delivery dates More efficient use of resources thus revealing capacity hidden by planning systems More timely use of materials, reducing inventories masking production issues Reduction of WIP Recognizing Visual Planning and Scheduling Systems The improved user interfaces in application software has increased productivity beyond simply automated planning and scheduling algorithms. As computer processing power has increased, more and more features have been added to planning and scheduling systems, causing increased similarity in how planning and scheduling applications look to the end user. One consequence is that a number of visual planning systems are advertised as scheduling systems, creating confusion within the user community, leading to end user disappointment after purchasing planning software thinking the package will solve scheduling problems. There are a few characteristics that the prospective software buyer can look for that will help separate a visual scheduling package from a visual planning system.
The first cue is distinguishing a visual planning system and a visual scheduling system is, well, visual. When the user opens the windows to view the allocation of work orders to resources, a bucketed planning system will display work as a vertical bar chart (see Figure 1) whereas a scheduling system will display work allocated to resources as a Gantt chart (see Figure 2), a format familiar to users of a project management tool such as Microsoft Project. Figure 1: Vertical Bar Chart of a Visual Planning System. Note how jobs are stacked in the time buckets.
Figure 2: Gantt Chart Display of a Visual Scheduling System. Vertical lines show daily break but do not delineate a time bucket The visual planning tool shows jobs in queues, but the jobs have no sequence with the buckets to efficiently manage the work. The Gantt chart in the visual scheduler shows a continuous timeline where jobs have specific sequence of execution, marked by specific start and end times, making it possible to generate dispatch lists that can be issued to operators on the floor, like sheet music to members of an orchestra, working towards a common goal. Another distinguishing characteristic that sets a scheduling system apart from a planning system is the wealth of data that a scheduling system can exploit to create effective, detailed production schedules. Examples of this data are: Resource and product specific constraints, such as a cutting blade when a certain product is run on a lathe Sequence dependent setups between products or product families, meaning the time to set up a machine is 30 minutes changing from Product A to Product B, but is 45 minutes from Product A to Product C Constraint calendars Detailed routings for products, including resource specific run times Unlimited scheduling horizons if needed Pegging between work orders or components to support complex assemblies, an example is that Work Orders A001, A002 and A005 must be completed before Work Order A010 can start Production and operation attributes to create efficient campaigns of work, such as diameter and metal type Slack time at the end of an operation, expressing a drying time after painting
Inter-operation time such as maximum time between operations, such as the metal coil must be placed in the annealing oven within 24 hours after the acid bath, otherwise corrosion will require another pass through the acid bath There is a capability that is a marked functional difference between a planning and scheduling system: the ability to either manually or automatically alter the sequencing of jobs on resources to maximize performance of production. In a system defined in time buckets, moving a job from one time bucket to another is unlikely to have an impact on the occurrence of subsequent operations in the following time buckets because of the gross granularity of the buckets. The natural question arises When is a time bucket small enough? The answer is: Neither the end user nor the software vendor really knows except in a few isolated cases, often limited to product line oriented, continuous (i.e. process) production. In a scheduling system, the lack of time buckets means true continuous relationship between operations without artificial time barriers, reflecting the same reality as that of the production floor. Conclusion Despite superficial similarities in user interface, and assertions by the vendors of planning systems as faux scheduling systems, the reality remains that a visual scheduling system will always render more precise sequences of actions for the production floor than those generated by visual capacity planning systems. For the prospective buyer of a visual scheduling system, one of the first characteristics to look for is whether the planning window is continuous or bucketed. If the time line is segmented into buckets, the application software is not a true scheduling system. Planning systems are attractive because of their simplicity but are limited to solving simple planning problems, and not scheduling problems. With today s computer technology, there is no excuse not to use a true scheduling system with the speed and visualization offered by a first rate visual scheduling system. About the Author Gregory Quinn is president and CEO of Quinn & Associates Inc, one of the most successful resellers of Preactor, the world s leading planning and scheduling solution. Mr. Quinn holds several degrees, including a dual Master s in Operations Research and Industrial Engineering from the Pennsylvania State University. He has taken decades of experience in different manufacturing companies to advance the benefits of scheduling for all industries.