Takt Time and How It Is Used

A basic concept of lean manufacturing is to schedule at only one point in the value-stream, pulling work from earlier processes and flowing product to the customer through all downstream processes. However, using a system for flow and pull presupposes that:

• The value stream has been designed to support the rate of customer demand (and buffer adjustment when necessary). In other words, manufacturing processes (cells) must be designed with enough equipment and labor to be able to produce to the required rate, suppliers must be capable of delivering the required volume and mix of component materials, and internal material delivery routes must be designed to operate at the same rate.

The point at which work is scheduled is called the pacemaker. Generally speaking a pacemaker is the point in the value stream that both

The pacemaker process controls the basic rate of production for a group of items with similar value-streams. Processes upstream of the pacemaker (components) are pulled from their source processes. Processes downstream of the pacemaker are handled on a first-in-first-out (FIFO) basis.

In a lean production environment, items that go through pacemakers would typically be the same as the ones that are master scheduled, regardless of whether the strategy for the parts is make-to-stock or make-to-order. Although a basic objective of lean manufacturing is to produce only to customer orders, in reality, there are times when producing to a small stock, typically as a way to buffer the variability of customer demand makes more sense than flexing labor and plant processes to day-to-day order variations.

In many lean environments, the master scheduled items would be produced to a finished goods buffer (a finished goods supermarket in lean parlance). In other lean environments, they would be produced only when the customer order has been received.

They might be produced from stocked component parts being held in a semi-finished stores area or they can be produced truly to order from raw material or purchased parts. But in every case, the master production schedule would be used to establish the future plan, the rate at which components will be pulled from supplying processes and outside suppliers, and the projected requirements that can be shared with suppliers for their planning processes. The finishing schedule, which might be the same as the master schedule or which might be different based on the mix of actual customer orders, would communicate what is actually to be produced in the pacemaker process.

The basic rate of manufacturing and the basis for the master production schedule and the finishing schedule is the takt time: literally the drumbeat for the process. Takt time expresses the sales rate and how fast the plant must produce a product in order to be perfectly synchronized with the customer. One unit every 23 seconds in a company making visors for automobiles, one unit per hour in an organization producing large pumps, or one unit every 2 weeks for a manufacturer of fighter jets would be practical examples of the takt time in real companies. Producing to takt time means producing to at the rate of the customer demand.

Takt time can be calculated from the customer demand and the working time:

Operational takt time expresses the anticipated build rate, taking into consideration finished goods inventory adjustments or work time adjustments like overtime. For the examples cited above, the operational takt time might be one unit every 23 seconds (no inventory adjustment or overtime), one unit per 1 hour 15 minutes (finished goods inventory is being lowered to reduce working capital, or one unit per every 1.5 weeks (overtime has been added and the build rate increased because of a subsequent planned shutdown).

Operational takt time can be calculated from the customer demand, buffer adjustments, and adjusted working time:

• Takt time is based on the total volume of all items being produced at the pacemaker process, not the quantity of any individual item. Takt time expresses the rate of production to match the rate of customer demand for the process. Another important and related number, operational takt time, expresses the rate of production to match the rate of customer demand and any needed finished goods inventory adjustments in aggregate and any adjustments to the work calendar needed to be able to meet the required volume. In a company which has implemented both resource planning and lean execution, takt time and operational takt time would be key outputs of the sales and operations planning process.

• When different products are produced in a single pacemaker, there will be an overall takt time, as well as a takt time for each product on the line. For example in an assembly cell in an engine factory, the takt time for the process might be one unit every 2 minutes 30 seconds while the takt time for the long block group of products is one every x min y sec, for short blocks is n min m sec, etc. The takt time by product will be used to calculate the appropriate mixed-model sequence of engine production.

• Takt time and operational takt time are calculated numbers, not a reflection of the capability of the process. There is no guarantee that what the customer wants can be supported by the capability of the supplying process.

• Over time, the objective is to get operational takt time to equal takt item, that is, to produce to customer demand. In the short term, an important activity associated with the pacemaker processes is to get the cycle time for each item to something less than or equal to the operational takt time, and in the longer term to get cycle time to where pure customer demand driven takt time can be supported.

• The industrial engineering activity which analyzes and adjusts the pacemaker process so as to align cycle time with takt time is called cell design or operator balancing. Other important design or planning activities related to takt time include those related to creating a mixed model schedule, creating the finishing schedule for the pacemaker process, material route design, and supermarket and loop sizing.

The finishing schedule is typically communicated to the plant using a visual control called the leveling box (heijunka box). The basic idea behind the heijunka box is to distribute the schedule evenly over time, ideally in small increments of time equal to the takt time, and then communicate the schedule to the plant in the same time increment. Typically it distributes the finishing schedule over a shift or a day, showing visually what is to be produced, when it should be produced (start and end time), and how much of each item in each time interval. It is also a principal driver of material route design for the plant since material deliveries to the pacemaker could be done in time increments based on the takt time driving the heijunka box.

However, if the takt time is very small, say 30 seconds, scheduling production or material routes based on time increments so small can be impractical. In this situation, a larger time increment called the pitch is used instead. The pitch for a process is some multiple of takt time, usually based on the number of units in a standard packaging unit, and typically between 10 minutes and 2 hours. It is sometimes called takt image because it preserves the drumbeat created by takt time, but makes it possible to support cell design and operator balancing (how many operators are needed in the cell and how should the work be distributed to those operators so as to meet takt time) as well as material route design (which parts should be delivered to each finishing cell, for what quantities and when).

In the next section we’ll cover a concept and calculation that is arguably as important to the beginning practitioner of lean as the takt time is to every serious practitioner.

Long setups are an impediment to flow. Longer setups imply larger lot sizes and lumpier demands, longer lead times, worse quality, more inventory and surges of work upstream. From a master scheduling and finishing scheduling perspective, the time to setup or changeover in a pacemaker can prevent a real mixed-model schedule and also stand in the way of implementing effective pull systems for upstream components.

Smaller lot sizes mean better flow, less overall inventory, and more responsiveness to changes in demand. So a logical question might be if smaller is good, then why not even smaller. Why not dictate lot sizes of one piece?

Of course if you’ve been around manufacturing for some time, and have some experience with the way the factory works, you know that there’s a limit to how small the lot size can be. If you reduce the lot size to too small a quantity, you will quickly eat up all your available capacity in changeovers. This is exactly what happened in some naive manufacturing companies who heard about all the benefits of one-piece flow and attempted to get there without doing the hard work to reduce manufacturing setup times.

The smallest possible lot size for a part in a process is a reflection of the production interval or EPE Interval. This interval defines the maximum frequency that you can produce each part without running out of capacity because of too many setups on all parts. It is an expression of how frequently each item produced in a process can actually be run without incurring a capacity penalty.

It answers the key question: how often can we run every part through the process: once a month, once a week, once a day, once an hour?

The essence of the EPE Interval calculation is to compute the capacity required for running all the items that go through a process, and use that to determine how much time is available for setup. The amount of time available for setup can be factored by the total setup time for producing every item at least once, and from this the maximum number of times you can set up every item and the order quantity for each one based on that frequency.

The benefits of running each item using the smallest valid interval include:

For the pacemaker process, reducing the setup and EPE Interval to less than the ship window has another potential benefit: the ability to meet small orders equally as well as large orders. With an EPE Interval of less than a day, any item can be run on demand. The goal of mixed-model scheduling is to get the EPE Interval to less than a day, shift or ship window.

The real question isn’t are smaller intervals better? The real questions are:

How small an interval now? and What do we have to do to get the interval smaller? and How fast can we get to a smaller interval?.

The basic calculation of the EPE Interval is this:

Think about it this way: if you determine how much time is left each day after producing enough of every part to cover the parts daily demand, this really tells you how much time you have in an average day to do setups. Say this is 3 hours per day. If your setups across all parts produced in the process total 3 hours, this means that you can run every part every day. If your setups across all parts total 6 hours, then you can at most run every part every two days (6/3=2).