This post is the second in a series on the science of pace of play management. Through his extensive research, Dr. Lou Riccio, FAIRWAYiQ's chief analytics officer, has developed actionable pace of play management techniques that are outlined in his "Pace of Play Bible." These proven techniques are based upon science and are the results of years of research - they are not hunches or guesses.
This week's topic in this series on pace of play management is "Factory Physics and Pace of Play". Here is the excerpt from Dr. Riccio's "Pace of Play Bible":
Factory Physics (definitions)
Before explaining the techniques used to study the problem, a brief explanation of the science of “factory physics” is appropriate since the approach of this work is to consider the golf course a factory producing rounds played by groups of golfers. The science provides a framework for analysis and brings with it some insights into problems and solutions for moving product (groups of golfers) through a factory (the golf course.) After that discussion, a review of previous studies is offered and then the analysis is explained.
To explain the concepts of factory physics, some definitions are needed. Definitions for “processes”, “operations”, “throughput time”, “cycle time”, “capacity”, “throughput rate”, “work-in-process”, and “bottlenecks” are offered to assist in the discussion of our analysis. Each is important to people who run factories and each has a special meaning and importance to the issue of pace of play and total rounds played.
All factories consist of processes which when provided with the right resources, produce finished products. Processes are made up of individual operations generally done in a series. Each operation adds something (adds value) to the product. In our golf analogy, each hole is an operation and each foursome is a work-in-process product which becomes a finished product when the group completes the 18th hole.
In factory physics, the time it takes one unit of product to go through the whole process, from start to finish, to go from raw input to finished product, is called Throughput Time. That would be equivalent to the time it takes one group to play their full round of golf.
Another time measure is the time between successive outputs, or completed products, of a factory. Stated another way, when one unit comes out of a factory, how long will it be until the next finished unit comes out? That is called Cycle Time. In the world of golf, that would be the time between successive groups finishing the 18th hole.
These two time indicators measure very different things. The distinction is important for the discussion later. The descriptions above apply these definitions to the whole process, all 18 holes, but they can also be applied to each operation (each hole) as well. The throughput time for a hole would be the time from when a group arrives at a tee until they complete the hole, including the time to play the hole plus the time waiting on the tee to get started. On the other hand, the cycle time for that hole would be found by standing at the green and starting a stopwatch as one group walks away from the green having finished the hole and stopping it when the next group walks away after finishing the hole.
Throughput time in a factory is comprised of three things: time in which the product is actually being worked on (operation or value-added time), waiting time (time from when a unit arrived at an operation until work began on it at that operation), and transportation time (time needed to move it to the next operation.) In golf, playing the hole (hitting and moving) is operation time, waiting until the group ahead has cleared the landing zone or green complex is waiting time, and the time walking to the next tee from the previous green is transportation time. For this analysis, walking between strokes is considered part of operation time since it is part of playing the hole.
In a well-designed factory, operation time is a large percentage of throughput time. Very little of the time is spent waiting or moving the product. For a golf course to be a well-designed factory, waiting time and time between holes should be minimized. The original rules of golf published by the Honourable Company of Edinburgh in 1744 said “ye shall tee your ball no further than one club length from the previous hole” – and they meant the hole itself. Golf architects should take note of that.
The amount of product that comes out of the factory per unit time (per hour or per day) is called the Throughput Rate or just Throughput (not to be confused with Throughput Time.) In golf, the number of groups who finish their rounds per hour or per day is the throughput. Throughput determines revenue.
Work In Process (WIP) is the sum of all product in the factory which is in production but not yet finished. In golf WIP is the number of golfers on the course at any one time. In the factory a certain amount of WIP is needed to keep the factory “humming” and to maximize Throughput Rate. However too much WIP clogs the factory and increases Throughput Time. Factory Physics has techniques to find the right amount of WIP to keep Throughput Rate high but Throughput Time low.
Those techniques calculate an optimal amount of WIP, or in golf, an optimal amount of groups on the course. Once reached, that optimal amount should be held constant, or at least not exceeded. Any increase over that amount will lead to longer Throughput Times. The optimal amount will keep revenue up yet not slow down play. In golf, once that amount has been reached, groups should only be allowed onto the course at the same rate as they come off. The tee time interval is the key to maintaining an optimal number of groups on the course.
The Capacity of a factory is the maximum amount of product it can finish per time period, per hour or per day. Whether the factory is making cars or Oreo cookies or tee shirts, there is a limit to how many it can produce per time unit. All production systems have a limit. How much is the limit? How do you calculate that limit? Good managers are always looking for ways to maximize their capacity, to get the most out of their factory.
Capacity differs from Throughput in that throughput is the actual amount produced while capacity is the most the factory can produce if everything is in order. Throughput can be less than or equal to capacity, but not more. If the orders for a product come in at a rate less than the factory can produce, throughput will be equal to the demand. If the demand is greater than capacity, throughput will equal capacity. A factory with a capacity of 100 units per hour may have a throughput less than 100 if machines break down or if the demand is less than 100 per hour. In golf, a course may be able to “produce” seven groups each hour, but if only six groups show up, the throughput would only be six even though its capacity was seven.
Capacity is a critical concept. Measuring it and understanding its implications is central to understanding how to maximize the rounds played in a day. The capacity of a factory is equal to the capacity of the lowest capacity operation. If a factory has several operations done in series, the one that can produce the least per hour is the one that determines how many units the entire factory can produce per hour. The lowest capacity operation is called the bottleneck. If a factory had five steps to produce a product, and the operations at four of those steps could produce 10 units per hour and the fifth could produce 8 units per hour, the factory's capacity would be 8 units per hour. Increasing the capacity of a non-bottleneck operation does nothing for increasing the capacity of the factory. Increasing one of the 10 units per hour operations to 11 units per hour will do nothing for the capacity of the factory. To increase the capacity of the factory, the capacity of the bottleneck has to be increased.
Capacity is important to a golf course because it determines how many groups can finish per hour (throughput rate) during daylight hours. For most of the day (once the course is full), capacity, not throughput time, determines how many groups can finish in an hour. As shall be shown later, only at the end of the day (as the sun goes down) does throughput time have an effect on the total number of rounds played.
Another important aspect of capacity is that it is ephemeral. An airplane which takes off with an empty seat cannot add another seat on the next plane to make up for the lost production. In golf when a group goes out with three golfers, the lost “productivity” cannot be made up by putting out a group of five later that day.
In a factory, bottlenecks are found in a number of ways. First as explained above, a bottleneck is the operation in a process which has the lowest capacity (lowest output per unit time.) Second, the bottleneck operation has the highest machine or worker utilization rate. It is always busy since it always has work to do. Third a bottleneck is the one operation with work-in-process inventory (things to work on) building up in front of it.
Of the three ways to identify a bottleneck, the last measure is the most obvious and easiest to use. On a golf course, 3 pars are usually the bottlenecks, the holes where groups queue up waiting to tee off. Interestingly, it takes less time to play a par 3 (throughput time) than a par 5, but the par three is the bottleneck. That condition will be discussed at length later in this paper. (Hint: we need to understand a hole's cycle time as well as its throughput time.)
In its simplest form, a factory that makes one product has one bottleneck. The bottleneck operation constrains the entire factory. In golf the bottleneck, like in the factory, may be one part of the course. However in golf the “product” itself may be the bottleneck in that one very slow group will slow everyone down. This other dimension adds a level of complexity not usually found in factory analysis. This complication will be analyzed later in the paper.
Of all the factors just defined, some are mathematically related and some are not. For example, capacity and cycle time are directly related. One is the inverse of the other. If the factory can produce 8 units per hour, on average the time between successive finished units will be 7.5 minutes, sixty minutes divided by 8. If you know one, you know the other. If you increase capacity, you reduce cycle time, and vice versa.
On the other hand, capacity and throughput time are only remotely related, related only by the time it takes to get through the bottleneck. For example in the five-operation factory discussed above, the capacity of the factory is equal to the capacity of the bottleneck, 8 units per hour. The other four operations have a capacity of 10 units per hour. The throughput time of this process is the sum of the times to complete all five operations. The time to get through the bottleneck operation (assuming no waiting time) would be 7.5 minutes, found by dividing 60 by 8 per hour. The time to get through each of the other four is 6 minutes found by dividing 60 by 10 per hour. Thus the total throughput time is 31.5 minutes.
If we now improved the capacity of one of the non-bottleneck operations to 12 per hour, the time to finish that operation would now be 5 minutes and overall throughput time would be improved. However, the factory can still only produce 8 per hour or one every 7.5 minutes, since the bottleneck has not changed. In that case, throughput time is improved but cycle time (and by definition throughput rate) is not changed. One lesson from this is changes may lead to the improvement in one measure of system performance but not others. In some cases it may lead to an improvement in one and a reduction in another. Understanding how all aspects work together, not just one or two, is critical to improving system performance.
One last point is important to this study. If the input rate to the system (factory unit orders per hour) is greater than the capacity of the bottleneck, work-in-process inventory will naturally build at the bottleneck. To insure WIP inventory doesn’t build up at the bottleneck, the capacity of the bottleneck has to match or slightly exceed input rate. In the example above, if factory orders were coming in faster than 8 per hour, WIP inventory would build in front of the bottleneck since it can handle only 8 per hour.
Regarding this last point, golf courses tend to send out golfers from the first tee as soon as the group ahead clears the first hole’s hitting zone. That time is often shorter than the cycle time of the bottleneck hole (typically a par three) thus resulting in a backup there. This problem will be discussed at length later.