Production Line Falling Behind? That Day My Director and I Had a Heated Argument Over Takt Time
That day I was really angry; I remember it was 2:30 PM, and the equipment utilization rate on the production line report had dropped again. The director rushed over, face dark, pointed at the report and said, "Ah Wei, tell me, why isn't production efficiency going up again? Did we even calculate our Takt Time correctly?" I said, "Director, I calculated the Takt Time very clearly; it's 12 seconds per unit." The director slammed his hand on the table: "Then why is the actual output still significantly short? Is my stopwatch broken?" To be honest, I was also furious at the time; the numbers were all calculated, but the actual operation just wasn't smooth. This is simply the frustrating reality every engineer encounters.
To Put It Simply, It's "Falling Behind"
Have you also encountered situations where the calculated standard work time looks great, but the actual production line just can't achieve it? The problem lies in that we often only see the efficiency of a "single machine," but forget that the entire "production line" is an interconnected system. Takt Time, to put it simply, is how often the customer needs us to deliver one product. It's not how fast your machine can run at its fastest, nor how fast you, the engineer, think it should run. It's what the market tells you: how often you need to produce one unit. If the customer needs 1000 pieces per day, and you work 10 hours a day (600 minutes), then your Takt Time is 600 minutes / 1000 pieces = 0.6 minutes/piece, or 36 seconds/piece.
So the key point is that Takt Time is determined by customer demand, not how fast your production line can produce. All your stations must complete their tasks within this Takt, in order to satisfy the customer.
In Reality, This Is How You "Identify the Bottleneck"
Finding the problem is actually not difficult.
- First, calculate your "Ideal Takt Time": As I mentioned earlier, use customer demand to calculate it. Assume customer monthly demand is 30,000 pieces, you work 20 days a month, 8 hours a day (480 minutes).
* Ideal Takt Time = 480 minutes/day * 60 seconds/minute / 1,500 pieces/day = 19.2 seconds/piece.
* This means you need to produce one product every 19.2 seconds on average.
- Next, measure the "Actual Cycle Time" for each station: You need to actually use a stopwatch to measure how long each machine or manual station takes from start to finish to complete one product. Don't trust the "average work time" displayed by the machine; that is often inflated.
- Compare and identify the "Bottleneck": Now compare your Ideal Takt Time (19.2 seconds) with the actual Cycle Time of each station.
* Station B (20 seconds) > Takt Time (19.2 seconds) – Problem!
* Station C (17 seconds) < Takt Time (19.2 seconds) – OK
* Station D (25 seconds) > Takt Time (19.2 seconds) – Big Problem!
In other words, if any station's Cycle Time is longer than your Takt Time, that is your bottleneck. It will slow down the entire production line, preventing your total output from ever meeting customer demand.
The Most Common Pitfall: Good Numbers, Poor Field Performance
Frankly, many times we engineers simply trust the numbers on reports too much. I once pushed machine parameters to the limit to increase the capacity of a process station, reducing the reported Cycle Time from 22 seconds to 19 seconds, perfectly matching the Takt Time (19 seconds). What was the result? The yield dropped from Cpk 1.33 to Cpk 1.08, and DPMO soared from 2700 to 6210.
To put it simply, this "pretty" Cycle Time was achieved at the expense of quality. The preceding stations produced faster, but due to an increase in defects, subsequent stations experienced more downtime, rework, and scrap, ultimately making the overall efficiency of the entire production line worse. This is the trap of "local optimization." You thought you solved one bottleneck, but you merely shifted the bottleneck elsewhere, or even created a bigger problem.
So the key point is, don't just look at single-point efficiency; look at the balance of the entire production line.
One Thing You Can Do Today
Grab a stopwatch and go measure the actual Cycle Time of the slowest station on your production line.