That day, when the CPK report came out, the room fell silent for three seconds, and I saw the PM's face turn green.
I recall an incident several years ago when one of our critical machines unexpectedly went down without warning. The process handled by that machine consistently produced wafers with excellent yield; a Cpk of 1.8 and above was standard. However, that day, the machine was down for six hours, bringing the entire production line to a halt. Once the machine was finally repaired and a batch of products was run, the yield report showed the Cpk dropping directly to 1.08, and DPMO soaring to 6210. The conference room instantly fell silent, and the PM's face turned green. That incident made me truly realize that equipment downtime isn't just about fixing the machine; the costs involved are far more alarming than we imagine.
Where's the Problem? Why Won't the Boss Listen to Us?
To put it plainly, many times when engineers try to explain the importance of "preventive maintenance" to the boss or management, we often fail to articulate the financial case clearly. We only say, "This machine is about to break down and needs maintenance," or "If we don't replace parts soon, it will eventually stop." What the boss hears, however, is "More money to spend, and more downtime." Both sides are simply talking past each other.
In fact, it's not that the boss doesn't want maintenance; it's that he can't quantify the cost of "not performing maintenance." He only sees the immediate maintenance fee but doesn't see the string of losses caused by downtime later on. In other words, we need to calculate that cost for him. Let him know that investing in this maintenance fee now actually saves several times that amount in the future.
How to Actually Do It? Break Down Downtime Costs
To convince the boss, you must use a language he understands: numbers. We need to break down the cost of equipment downtime into several components:
- Direct Losses: This is the simplest. How long the machine is down, how long the production line is stopped, directly equates to how much production capacity is lost. For example, if one machine produces 10 wafers per hour, and each wafer is worth $500, then one hour of downtime means a loss of $5,000. If it's down for six hours, $30,000 is gone.
- Indirect Losses (Yield Impact): This is the most painful. As I just mentioned, that time the Cpk dropped to 1.08, DPMO to 6210. What does this mean for the defect rate in that batch of wafers? Suppose the usual defect rate is 0.1%, but that time it surged to 1% due to instability. Each extra 0.9% of scrapped product represents money. You need to multiply the total value of that batch of wafers by the increased defect rate to get your yield loss.
- Labor Costs: As soon as equipment stops, PMs, PEs, TEs, and even R&D personnel have to rush in to find solutions. Aren't these people's salaries a cost? If their time could be spent on other more valuable tasks, wouldn't that also create more benefits?
- Customer Trust: This is the hardest to quantify but also the most fatal. Long-term unstable delivery times and fluctuating yields will make customers leave. Once a customer transfers orders to someone else, it becomes difficult to win them back.
So the key point is, when you lay out all these numbers, you'll discover a startling fact: the losses from a single unexpected downtime incident can far exceed your entire year's preventive maintenance costs.
The Most Common Trap: Thinking It's Fixed Once Repaired
Honestly, I used to make this mistake frequently. When a machine stopped, I'd quickly find someone to fix it, and once it was repaired and the production line was running, I felt the task was complete. But in reality, the true damage had just begun.
Did you think it's all fine once fixed? Wrong! Many times, that unexpected downtime acts like an invisible landmine, buried within subsequent batches of products. At best, it causes yield fluctuations; at worst, direct scrap. Because a machine stops unexpectedly, many parameters can drift, and it takes time to stabilize after a restart. During this unstable period, the quality of products produced is questionable. You might even have to spend more time and labor on traceability and verification. These hidden costs are the easiest to overlook.
Frankly, many times we equipment engineers are overwhelmed by these annoying issues. The boss only sees the immediate repair costs but fails to see the cascading effects behind them. Therefore, we must proactively present these numbers and potential risks clearly to him.
One Thing You Can Do Today
Go back and calculate the "yield loss" your department incurred over the past year due to unexpected downtime.