Pull Scheduling / Just In Time
Pull Scheduling is generally synonymous with Just-In-Time (JIT) inventory management - a system of production scheduling that pulls product through the value-stream in a continuous flow rather than pushing it through in batches.
Prior to the development of JIT at Toyota, part production was generally scheduled centrally in batches for all upstream and downstream processes. The size of the batch was a function of the set-up time of the producing equipment. Batches were delivered to downstream processes whether they needed them or not, commonly resulting in excess inventory.
The pull system was developed to eliminate overproduction, and hence in-process inventory - seen a manifestation of waste.
In addition to the wasted capital tied up in excess inventory, other forms of waste include excess production resources, excess storage space, unnecessary movement of inventory, unnecessary inventory transactions, delayed feedback on quality problems, lengthened cycle times, and poor communication from visual obstruction of the workplace (you can't measure what you can't see).
Parts on Demand
In a JIT system, downstream processes withdraw parts as needed from upstream processes in small lots (ideally lot sizes of one). The mechanism to call for parts is a Kanban - a card specifying how many parts are to be moved.
A JIT system is tightly linked, with little or no buffer inventory between workstations, and therefore requires several interdependent conditions in order to function:
The rate of production must be balanced throughout the system based on the Takt time. Takt time is defined as the available production time for a period divided by the average customer demand for that period.
For example, if a two-shift operation has a net production time of 4,200 minutes per week (7 hours per shift after subtracting lunch and break time x 2 x 5 days), and the average customer demand per week is 4,500 units, then the Takt time is 4,200/4,500, or 0.93 minutes. A unit must be produced every 0.93 minutes in order to satisfy demand.
The actual production rate is called the Cycle Time. Production cycle time that is lower than the Takt time will result in excess inventory. The goal is to balance all processes to run at this same rate so that there is no in-process inventory accumulation between process stages.
Think of line balancing in terms of the typical home washer and dryer. The washer usually has a cycle time that is much faster than the dryer. If you continue to load and unload the washer at the end of it's cycle, inventory of wet clothes piles up ahead of the dryer.
The greatest efficiency and lowest cost is achieved when production is stable. Otherwise, waste would be incurred to always have capacity for peaks. Stability in overall production is achieved by mixing models in final assembly according to demand. Marketing efforts are also made to stabilize demand.
If set-ups or changeovers are lengthy, it is impossible to run small lots of parts. For example, large stamping dies commonly took hours to change until the development of quick die change methods (Single Minute Exchange of Die, or SMED).
Long changeovers necessitate cycle times well below the Takt time, and hence a buildup of inventory to supply production while the die is being changed. Information on Quick Changeovers
Reduced Lot Size
Reduced lot sizes go hand-in-hand with quick set-ups. The smaller the lot size, the better the flow through the process. Small lots can be passed continually between processes, allowing layout changes to link processes into cells, and reducing inventory and the space it requires. Quality and processing problems are quickly exposed and problem resolution is forced because there is no buffer inventory to hide the inefficiencies. With large batches, all sense of Takt time is lost on the shop floor. Information on One Piece Continuous Flow
In order to achieve a balanced work flow, cycle time equal to Takt time, and high quality, work is standardized at all operations for optimum efficiency and consistency.
Variability is removed by eliminating "adjustment", searching for tools and parts, and awkward ergonomics. The standard work sheet contains the cycle time, work sequence, and standard inventory, and is well documented and displayed at the workstation. Information on One Piece Continuous Flow
Continuous Improvement (Kaizen)
Continuous improvement, or Kaizen, is a process that surrounds a JIT system, most of which are established by a series of ongoing incremental improvements by teams focusing on better ways to operate - typically with little capital. A lean system exposes problems and demands resolution. Information on Kaizen Events
In the words of Taiichi Ohno, creator of the Toyota Production System, "autonomation refers to automating a process to include inspection. Human attention is necessary only when a defect is detected (the machine will stop and not continue until the problem is solved)."
This automation with a human touch only requires human involvement when there is a problem to be solved, so one operator can tend to multiple machines. See the Toolbox tutorial on Error-Proofing for related information.
Reduction of inventory between processes and reduction in lot sizes allows a reconfiguring of equipment geography, typically into cells organized by product family.
This is quite different than a job-shop orientation, where similar machines or processes are located together in departments and the product follows a tangled path from department to department. Production cells often utilized smaller, dedicated machines or tools to eliminate set-ups entirely.
Reliable equipment is crucial to a JIT system as there are no buffer inventories to cushion production downtime. Total Productive Maintenance is used to ensure high equipment uptime and effectiveness.
Processes that produce defects are not predictable and therefore create scheduling problems. A rule of JIT is that no defective product is ever passed on to a downstream operation.
If cycle times are appropriately determined (equal to Takt time), then the production of defects in any process causes production shortfall and hence immediate visibility. In this sense, JIT systems are often described by the analogy of lowering the water level in a stream to expose the rocks.
Process Playground is an online application designed to help you build and simulate your process models.
With the tools and automated features in Process Playground, you can quickly map your existing process flow! You then embed decision rules and data inputs to predict how your process configuration will perform. Simulation of the process helps you to visualize and quantify the effects of your potential process improvements. Learn more about this process mapping and simulation tool.