Maximizing Profitability Part 12

One of the key points to remember regarding the implementation of Drum-Buffer-Rope is that it will not be sustainable unless traditional performance metrics like operator efficiency, machine utilization and incentive systems are either abandoned or radically changed to be compatible with the principles of TOC. So, if these aren’t the correct metrics to track in non-constraints, then what metrics should we track. In my last posting we discussed buffer status….doesn’t it make sense that the non-constraint metrics should be measuring, in some form, how we are doing with respect to buffer management?

Remember, the properties of good performance metrics are:

• They must be objective, precisely defined, and quantifiable

• They must be well within the control of the people or process that is being measured

• They must stimulate the right behaviors

So, what behaviors are we trying to encourage in non-constraint operations? Think for a moment about the function of a non-constraint. First, a non-constraint in front of a constraint can never let the constraint starve. If it did, valuable throughput would be lost to the system. By the same token, a non-constraint downstream from the constraint can never be permitted to produce scraps because that too would be throughput lost to the system. Secondly, you know that protective inventory in front of the constraint simply adds unnecessary cost to the organization. Third, you know that the constraint operation sets the pace for all other resources, so schedule compliance in a non-constraint becomes important. So based upon these three functions, there are four generic metrics that should be considered:

• Workstation availability: The percentage of time the non-constraint was available to make product.
• Yield: The percentage of “good” product produced for the constraint operation (or for downstream operations supplying the shipping buffer).
• On-Time Delivery to the next operation:  The percentage of compliance to schedule.
• Protective Buffer (parts and/or time) in front of the constraint, assembly, and shipping too much (greater than 100 percent. Percentage of protective buffer too little (less than 100 percent)

The final decision on which metrics to use for a non-constraint is situation dependent, but the decision should be based upon:

• Never starving a constraint operation (or shipping buffer)

• Impact of too little or too much inventory in the system

• Schedule compliance

Let’s finish the discussion on the UIC by looking at steps 4a, 4b and 4c. In these three steps we are concerned with planning how to elevate a constraint if it is necessary, actually elevating it and then implementing protective controls to sustain the gains we have made.

With any luck, by following the steps I’ve presented, you will have increased the capacity of your constraint to meet the demands of the marketplace. If this is the case, then other, more extreme actions must be taken which usually involve spending some money. You have eliminated much of the waste and reduced variation, both of which have had a positive impact on cycle time and throughput, but you’re not quite there yet as far as producing enough product.

Goldratt’s fourth step tells us, if necessary, elevate the constraint. Increasing the capacity of the constraint can be done in a variety of ways like, using overtime, adding resources, purchasing additional equipment, etc. One thing you must keep in mind when you are elevating the constraint is what will happen when you are successful with this elevation. Remember back in Step 1, we said that you should identify both the current and next constraint? We did this for a reason. Suppose you have decided that in order to elevate the constraint, you must purchase a new piece of equipment. Your justification should only demonstrate the throughput improvement up to the limit of the next constraint. That is, if your current constraint is currently producing 5 parts per hour and you purchasing a new machine will double that to 10 parts per hour, that improvement is only correct if the next slowest resource is at 10 parts (or above) per hour. If, for example, the next constraint in the process is only producing 7 parts per hour, then you really can only claim a gain in throughput of 2 parts per hour for the new equipment. All I am saying is that make sure you consider the total process when making you decision on how to elevate the current constraint.

The final step in the UIC is to Implement Controls to Sustain the Gains. Sustaining the gains is the hallmark of great organizations, so how do we do this. Of course, if you have chosen the right performance metrics and you’re tracking them, this is one way, but is it the best way? One of the most effective tools to protect and preserve your accomplishments is by using a simple process audit. A typical process audit is a series of questions asked to the line leader and/or supervisor to demonstrate the status of the process. These questions should be focused upon how elements of your new process are working and to demonstrate that they are being used as intended. For example, suppose part of your plan involved implementing a control chart. Your questions would be focused on whether or not it’s being used; if out-of-control data points are being investigated, etc. You would of course then check the status of the control chart to verify the responses. These audits should not be planned in advance, but rather be done in a random time frame. One of the reasons audits of this nature fail to add value is that many times they are announced in advance. Anyone can look good for a day if they are given enough advanced notice. I also advocate having leadership perform some or all of these audits simply because it adds credibility to them. If leadership thinks they are important, then everyone else will too. In addition, I recommend scoring the audit by % compliance and posting the audit score as demonstrated in the  figure below.

You have just completed one revolution of the Ultimate Improvement Cycle and things should be greatly improved. Your throughput has increased, your cycle times are reduced, your quality is better, there is less variation and uncertainty and your bottom line is much better. Don’t stop here….move on to your next revolution and start the process all over again. This is your new Process of On-Going Improvement…your POOGI. Good luck!