Insourcing, outsourcing, offshoring, automation, and other strategies are used to create, expand, and manage global productive capacity. Organizations now create capacity to satisfy expected demand anywhere, anytime, and based on service agreements at the lowest possible investment cost and risk. Information, technology, virtual teams, and new process structures are critical components of capacity around the world in recent years. In these new systems, capacity has some physical presence, but it also has an increasingly greater virtual content. An example is subject matter expertise accessible anywhere in the world at any time, as opposed to creating a local team that is off schedule part of the day.
Capacity is the ability of a transformation system to produce goods or services according to a schedule at an agreed-upon time, location, and quantity. It is a measure of a system’s ability to efficiently transform inputs, represented by materials, labor, machines, information, energy, etc., into outputs, represented by products, services, or information. Capacity is also measured on a fixed time basis under predefined conditions related to system availability and service-level agreements. Based on predefined demand, delivery promises can be met if the predefined conditions exist. For a given system, its capacity can be modified to satisfy demand and delivery promises dependent on its design limitations. Capacity can also be thought of as a combination of labor, materials, and capital brought together to do work. Capacity will be reduced if the effectiveness and efficiency of resource utilization are below standard. It can be considered from different perspectives such as available, temporarily stored, or made available at a future date by planning. Processes can be designed to have each of these three forms of capacity in various parts. As an example, in some systems, 100% of the capacity is on-site and available. In other systems, it can be made available as needed through supplier agreements, leasing equipment, temporary labor, self-service systems, and related strategies that create capacity based on demand.
A system’s capacity is dynamic. It changes based on the types of products or services flowing through a given process at a given point in time. Variations in the efficient use of labor and capital or the impact from external factors are additional causes for dynamic fluctuations in a system’s available capacity. Capacity can also be defined as design capacity, available capacity, and actual capacity. Design capacity is the throughput of a process or system if all labor and capital are working at optimal levels. Available capacity is the system’s design capacity minus expected or planned loses in operational efficiency because of technology or process constraints. These are caused by known time lost due to inefficient job setups, inspection of work, processing of work (e.g., moving materials or information), and waiting. The third type is actual capacity and is defined as the available capacity minus time lost due to unexpected events that negatively impact the throughput. Examples include scrap, rework, schedule variations, lack of resources, and other unexpected causes.
Recall that a system’s throughput is the measure of available capacity at its bottleneck resource and, depending on operational variation within the system, its capacity-constrained resources, which may periodically become system bottlenecks. This requires increasing bottleneck capacity. Operational capacity at non-bottlenecks is also increased by simplifying, standardizing, and mistake-proofing processes and using other Lean tools and methods. Examples discussed in Chapter 6 included establishing a takt time to utilize capacity only if needed to satisfy external customer demand, operational balancing of work, the use of transfer batches, the application of mixed-model scheduling, waste elimination to reduce rework and scrap, and deployment of pull systems. Quality improvement tools and methods, such as Six Sigma or Design for Six Sigma, can also be used to reduce process variation to improve yields.
Economies of scale also increase capacity. If a system is operating below its available capacity, indirect costs must be allocated across a smaller throughput quantity. This increases unit costs. As a system’s throughput increases, unit costs decrease because of operational efficiencies. A model is shown in Figure 13.3. As an example, there are fewer required job setups and changeovers from one product to another. This enables greater
Economy of scale.
standardization of production methods. The result is an increase in the system’s utilization and indirect costs can be allocated over larger production quantities, which reduces per-unit cost. The goal is to achieve economies of scale but also to be adaptable in meeting changes in demand though design standardization and process improvements such as mixed model scheduling.
At this point, we should distinguish between resource utilization versus activation. This is best understood in the context of the bottleneck resource. When upstream operations feed a downstream bottleneck, they should be activated only when the bottleneck needs material or, in service industries, information for production. Recall that a non-bottleneck resource should be balanced to the system bottleneck to avoid excess inventory and other issues. In other words, the utilization of non-bottleneck resources should match the bottleneck’s activation and likely should not be utilized 100% of the time. A complicating factor in this scenario is that other products may be starved for materials and components if a production schedule changes if non-bottleneck resources are utilized unnecessarily. If additional system capacity is needed, then it should be added at the bottleneck. Examples include adding low-cost redundant machines, allocating multiskilled workers, or improving yield and reducing operational cycle times at the bottleneck.
As workers learn better ways to do work and engineers optimize the process technology, the lowest cost curve shifts downward, reflecting decreasing per-unit costs even at lower system throughput volumes. How is this possible? Learning curve theory assumes that an improved understanding of a process is gained though experience and contributes to higher operational efficiencies. These in turn contribute to higher available system capacity. Historically, less than optimal and more expensive product and service process designs drive supporting process and infrastructure design and cost to create go-to-market capacity. If markets become differentiated and move toward niches based on customer preferences, it becomes increasingly difficult to gain economies of scale to lower costs without design and process changes.
Adapting strategies for good design, organizational structure, automation, virtualization, and other methods must be used to lower costs. Technology deployment enables an increasingly higher percentage of virtualization to allow information to be transmitted quickly to do work. This eliminates the need for building physical infrastructure. It also creates supply chains where capacity can be inexpensively increased to match demand. In industries requiring production infrastructure such as oil, gas, utilities, and manufacturing, product design and supporting process design drive cost and system flexibility. Economies of scale are still applicable in these industries, as are the many tools and methods discussed in previous chapters.
Entire industries have moved away from a heavy reliance on physical infrastructure, although infrastructure is still integral to operations, just to a lesser degree. The telecommunications industry is an example where technology shifted the economy-of-scale model to lower costs as system throughputs dramatically increased. As an example, new cell phone technologies were set up using satellite technology without laying down expensive cabling in many regions of the world. There is still infrastructure, but not at the previous level.