Multiple Machine Operation Revisited

Multiple machine operation has been the subject of several recent Tech Talk columns (January through May of 2008). I have received many comments, and I want to thank everyone who responded. The series states that successful applications for having one operator run multiple machines do not just happen—a successful application is engineered.

According to Robert Crowley of New Age Technologies, “The issue of how to best use an operator’s time is an issue in a lot of shops.” While deburring parts is a necessary function, is not necessarily a smart operator assignment, he says, especially when the operator is having problems keeping up with the machines. “The assignment of this task tends to be an easy fix whenever processing and part programming is not up to standard. It is far better to process and program jobs so that the required operator time per part, exclusive of the loading time, is kept at a minimum,” Mr. Crowley says.

I agree. Cutting tools and programming methods can eliminate almost any burr from a workpiece. In some cases, deburring can be done without increasing program execution time. Consider, for example, chamfering a shoulder on a workpiece machined on a CNC turning center. If a tiny radius (0.002 inch to 0.005 inch) is programmed at each corner of the chamfer—something very easy to do with a CAM system—the sharp corners on the chamfer will be nicely deburred in cycle with no increase in program execution time.

In other cases, special deburring tools will be required. While these tools do add to program execution time, they also ensure workpiece consistency and eliminate a time-consuming and sometimes inconsistent operator task. Indeed, if the machine sits idle while an operator deburrs a workpiece (as is often the case when an operator cannot keep up with the machines), having the CNC machine perform the deburring operations will free the operator to do other things. In marginal cases, this can make the difference between a feasible and an infeasible application.

“The same holds true for the entry of quality control data,” Mr. Crowley says. He adds that all current quality control systems have an objective to simplify operator entries by making it easy to input quality and standard deviation data. Part probing eliminates the operator from the picture altogether, he says.

This also relates to appropriately engineering an application—ensuring that the operator will have enough time to perform the required tasks during the production run while minimizing or eliminating downtime for any of the machines.

Not all of the comments I received were positive. One commenter says, “I understand that being specific about a given application is difficult. And, I do understand there is a point of diminishing returns for the one-operator/two-machines issue. However, several of my shop people who have read your article came to the conclusion that one-operator/two-machines in our facility is just not right. Obviously, they misunderstood the intent. I guess my point is that some people will read at the first level of understanding—and [they] don’t understand the real point. So, now I’m running interference and trying to help them understand the point you were really trying to make.”

Admittedly, the points make in this series can be misinterpreted. My intention, first and foremost, is to provide a way to gage whether having one person run multiple machines is feasible from a cost standpoint. Management should be able (and willing) to provide justification to people questioning this feasibility. As long as the cost for interference (downtime when one machine sits idle waiting for the operator to do something on the other machine) is consistently less per hour than the hourly cost of one operator (wages plus benefits), it makes sense to have one operator run two machines. Assuming a company knows its costs, monitoring the machines for a few jobs will provide some pretty good data for this evaluation.

Another commenter writes: “I was hoping you would address the issue of operator utilization differently. Operators are often assigned to machines with too little consideration for skill level.” This commenter goes on to say that as machines become free, they are assigned to the first available operator with no consideration for whether the operator can efficiently run them. “Inexperienced operators often result in wasted time,” the commenter says, “and in this case, even a ‘well-engineered’ application will fail.”

Again, I agree. Operator skill level is a topic that I do not address in the series. Many companies are struggling to find and hire people to run their CNC machine tools. It is not uncommon for newcomers to be inexperienced in the shop environment, which is all the more reason for management to appropriately design applications—taking into consideration the lowest skill level of the people involved.

The commenter continues, “We assign new operators to work with experienced operators until they pick up the needed skill and experience. We try not to assign time limits because each person is different. It works well for new people to have a ‘mentor’ within the company. Over the long haul, it has encouraged some very good working relationships.”

Again, thanks to everyone who contributed comments.