When the QC supervisor applies SPC to family life: An experiment on process control of "time to get home"
My friend Xiao Yang's "SOS game" started with the A3 paper at the entrance.
As soon as I rang the doorbell of his house after work on Friday, the door suddenly flew open. Xiao Yang stood at the door with one button of his shirt collar fastened incorrectly, and his eyes seemed to be saying "Help me!" "Brother, you're finally here!" There was a handwritten control chart pasted on the refrigerator in the entrance hall. The horizontal axis was labeled "Monday to Friday", and the vertical axis was "Time of arrival home". Two lines were marked with a red marker - the upper limit was 21:00 (oh no, it was 7 p.m.) and the lower limit was 18:00 (6 p.m.). The dots for each day were circled with a blue gel pen. The dots for the past seven days were connected to form an upward - sloping curve, and the last two dots directly pierced through the red line.
"This was painted by my wife," Xiao Yang rubbed his hands, his voice tinged with grievance. "She's the QC supervisor at the American-owned auto parts factory. Usually at the company, she manages the SPC on the production line – that's statistical process control. She specifically keeps an eye on the fluctuations in the dimensions of the parts. Now, well, she's applying the same approach to me!"
I leaned in to look at the control chart. There were also small characters written in the bottom right corner: "The criteria for identifying abnormalities in Shewhart control charts apply." Well, this wasn't just a randomly drawn "supervision form"; it was a "process monitoring tool" used by professionals.
I. Why SPC? A QC supervisor's "favorite tool"
Xiao Yang's wife, Sister Lin, poked her head out of the kitchen. There was tomato juice on her apron, and she was still holding a spatula in her hand. "Don't listen to his nonsense. I'm just showing my concern for him in the most familiar way!" She wiped her hands, moved a chair over and sat down. Then she took out a copy of the "SPC Manual" from her bag - the kind jointly compiled by the three major American automobile companies (Ford, General Motors, and Chrysler). The cover was all wrinkled from being flipped through.
"Do you know what I do in the company?" Sister Lin opened the manual and pointed to a certain page. "Our production line makes automobile pistons. The diameter tolerance is ±0.02mm. We need to measure 50 samples every day and draw an X-Rm control chart (individuals - moving range chart). If there are 7 consecutive rising points, it means the cutting tool is wearing out and we have to change it immediately. Otherwise, once the parts exceed the tolerance, the whole batch will have to be scrapped."
She pointed at the control chart on the refrigerator and said, "The same goes for Xiao Yang's 'process of going home'. Under normal circumstances, he gets off work at 5 o'clock, arrives at the parking lot at 5:15, and it takes a 45 - minute drive. He gets home around 6 o'clock. This is a stable process, and the fluctuation should be within 15 minutes. But in the past seven days, he has been 10 minutes late every day, with a continuous upward trend. This is not a random fluctuation. There is a systematic cause!"
Oh, I see. Sister Lin wasn't checking up on her husband. Instead, she regarded her husband's "time of getting home" as the "process output" she cared about the most. Just like how she monitors the piston size at the company, she uses SPC to detect trends in advance and prevent the occurrence of "non - conforming products" like "staying out all night".
II. The logic for detecting abnormalities of the control chart: It is not about catching the culprit red - handed but preventing problems before they occur
Sister Lin put down the manual and counted on her fingers the three criteria for judging abnormality in the control chart, which was the "gold standard" she used on the production line every day.
1. Out of bounds (beyond the control limit): For example, Xiao Yang arrived home after 7 p.m. in the past two days. This indicates that the process suddenly "lost control", just like a sudden power outage on the production line and the part dimensions suddenly exceeded the tolerance.
2. Seven consecutive points rising/falling/on one side of the center line: This is an abnormal trend, indicating that there is a continuously acting cause—such as the wear of the cutting tools on the production line, or Xiao Yang encountering the same traffic jam every day;
3. Non-random patterns: For example, it's always late every Friday, or it fluctuates once every two days. This is a "periodic anomaly". For example, one has to work overtime every Friday or work night shifts every other day.
"Xiaoyang currently meets the first two criteria," Sister Lin glared at Xiaoyang. "This is not 'occasionally late'; it's 'constantly late'. If this were the production line, I'd have issued an 'exception handling form' long ago!"
Xiao Yang shrank his neck and said, "I've already told her that they're repairing the outer ring road, but she doesn't believe me..."
III. Find the causes using the cause-and-effect diagram: Logical transfer from the "production line" to the "way home"
Sister Lin took out a piece of white paper and drew a fishbone diagram (cause-and-effect diagram) – this is one of the old seven QC tools. She uses it every day in the company to analyze the causes of defective products.
Person (Xiaoyang himself): Is it about delaying leaving work, like slacking off and using the phone? ——Xiaoyang said, "I turn off my computer promptly at 5 o'clock every day."
Machine (tool): Is the car broken? —— Xiao Yang bought his car last year and it has just been serviced.
Material (raw material): Is it out of oil? —— Xiao Yang refuels it every morning.
Method: Has the route changed? —— Xiao Yang said, "We always take the outer ring road. There's only this one way from the company to home."
Environment: Is there a traffic jam? —— Xiao Yang said, "They are doing road construction on the Pudong section of the Outer Ring Road. There are only two lanes left at the Yanggao Road Interchange. It was congested for an hour yesterday."
Sister Lin stared at the branch of the "ring" and frowned. "I thought he was making up an excuse..."
"I just passed by there last week," I quickly testified for Xiao Yang. "The maintenance of the outer ring road will last for a month. The westbound lanes have been reduced by half, and it gets extremely congested during peak hours."
Xiao Yang grabbed onto it as if it were a life-saving straw and said, "Look! I'm not lying to you!"
IV. Control Limits vs. Specification Limits: A "Minor Slip" of a Professional
Suddenly, Sister Lin laughed and pointed at the control chart, saying, "Actually, I made a small mistake. The range of 1 hour should be the 'Specification Limit', not the 'Control Limit'."
She opened the SPC Manual and pointed to a passage, saying, "The specification limit (Spec Limit) is the 'range of requirements'. For example, 'You should get home between 6 o'clock and 7 o'clock'. The control limit (Control Limit) is the 'fluctuation range when the process is stable'. For example, under normal circumstances, Xiao Yang's home - arrival time fluctuates within 15 minutes. Then the control limit should be from 6:15 to 6:45, which is narrower than the specification limit."
"I just took the easy way out before and used the specification limits as the control limits," Sister Lin said somewhat sheepishly. "Actually, the control limits should be calculated using historical data. For example, for Xiao Yang's arrival times in the past three months, we calculate the average and the moving range, and then use the formula UCL = X̄ + 2.66R̄, LCL = X̄ - 2.66R̄ (Shewhart control limit formula). Only in this way can it be accurate."
Oh, I see. Sister Lin's "control chart" wasn't "deliberately holding Young Yang back." She simplified a professional tool and applied it at home. After all, who would calculate such complex formulas for their loved one?
V. Adjust control limits: The core of SPC is "adapt to changes"
"What should we do now?" Xiao Yang asked.
"For anomalies caused by special reasons, the process needs to be adjusted," Sister Lin closed the manual. "The road construction on the outer ring road is a'special reason' – it's not Xiao Yang's fault; it's the change in the environment. So the control limits have to be changed."
She picked up the marker and changed the upper limit to 20:00 (8 p.m.), while keeping the lower limit at 18:00. "We'll adjust the control limits back to the original ones after the road repair is finished."
Xiao Yang let out a sigh of relief, picked up the cup and took a sip of tea. "You should have said so earlier..."
Sister Lin gave him a pat and said, "Aren't I just afraid that you'll be like when you were a kid and forget to go home while reading picture books?"
Oh, so there's such a story. When Xiao Yang was a child, he always went to the bookstall to read *Romance of the Three Kingdoms* after school. He forgot the time several times. Sister Lin (a little girl from the neighboring family at that time) went looking for him with her mother. When they found him, he was squatting on the ground, flipping through the books, with his ears red from being tugged.
VI. The essence of SPC: It's not monitoring, but "prevention love"
At dinner, Sister Lin served tomato and beef brisket. It tasted great, just as precise as her work in quality control at the company. Holding a wine glass, she said, "Actually, I didn't want to investigate him. I was just afraid something might happen to him."
"Look," she pointed at the control chart, "if I hadn't drawn this chart, and he was 10 minutes late every day, I might only complain, 'You're late again.' But with SPC, I can see the 'trend' — oh, it turns out he isn't doing it on purpose. It's because of the traffic jam."
Suddenly, I realized that SPC is not a "monitoring tool" but a "preventive tool". Just like Sister Lin uses it in the company to replace the cutting tools in advance, thus avoiding the scrapping of a whole batch of parts; and at home, she uses it to detect "traffic jams" in advance, thus avoiding misunderstandings between husband and wife.
Xiao Yang picked up a piece of beef brisket and said with a smile, "So my wife is caring about me in the way she's most familiar with."
Finally: From the production line to daily life, the humanistic core of SPC
When leaving Xiao Yang's house, Sister Lin turned over the control chart. There was a line of words written on the back: "Only when the process is stable can love be stable."
In fact, the essence of SPC has never been "data", but "understanding the process". Whether it's the parts on the production line or your lover's way home, as long as you regard it as a "stable process", use data to identify trends and logical analysis to find out the reasons, you can solve problems in advance.
As Sister Lin said, "Quality control is not about 'catching mistakes'; it's about'making the process more stable' — and the same goes for love."
That night, Xiaoyang walked me downstairs. Feeling the movie magazine (with a free DVD) in his pocket, he said, "Brother, I'll buy her a bunch of flowers tomorrow to thank her for the control chart."
I smiled and patted him on the shoulder. "What you should be grateful for is that she used her professionalism to turn love into a measurable'stable process'."
The aroma of tomato and beef brisket wafts in the air. There's still a traffic jam on the outer ring road in the distance. But at Xiao Yang's place at this moment, everything is going smoothly, and so is the love.
Manufacturing logic of complex products: Debugging is the inevitable path to stability
The complexity of system-level products like automobiles far exceeds the imagination of ordinary people. A car consists of 10,000 - 20,000 components and involves four major processes: stamping, welding, painting, and final assembly. At each step, the problem of "synergistic error" needs to be solved. For example, for the body frame in the welding workshop, the precision must be controlled within 0.5 millimeters; otherwise, the gap between the car door and the body will vary during final assembly. When hoisting the engine on the final assembly line, the coaxiality error with the transmission input shaft must be controlled within 0.1 millimeters; otherwise, it will lead to jerky gear shifting or even mechanical damage. These problems cannot be solved by design drawings alone. Small-batch trial production must be carried out for repeated debugging, such as adjusting the pressure parameters of welding robots, calibrating the positioning coordinates of engine hoisting, and testing the compatibility of electronic systems (car infotainment systems and sensors). The essence of this process is to turn "uncertain processes" into "repeatable standards." When the process capability index (CPK) ≥ 1.33, it means that this production line can stably output products that meet the quality requirements. This is the "process capability" in the manufacturing industry. In other words, the stability of complex products is never "innate" but "tuned out."
Pain points of single-piece production: Debugging costs crush controllability
However, the production of "single-piece products" completely breaks this logic. For example, when customizing an exclusive supercar, the customer requests that "the chassis tuning should match their own body weight of 80 kilograms", "the seat wrapping should fit the curve of a waist circumference of 85 centimeters", and "the exhaust sound should be like 'the beginning of Mozart's Symphony No. 40'". There is no ready-made parameter library for these requirements, and each item has to be "trial-and-error from scratch": the hardness of the chassis springs needs to be changed 3 - 5 times, the density of the seat foam needs to be sampled 5 - 8 times, and the structure of the exhaust muffler needs to be modified more than 10 times. More importantly, the cost of these adjustments cannot be spread through mass production: the adjustment cost for a single piece is often 5 - 10 times that of a mass-produced vehicle (for example, if the cost of adjusting a mass-produced vehicle is 10,000 yuan, it may cost 50,000 - 100,000 yuan for a single piece). Even if you spend so much money, you can't guarantee that the result will meet the expectations 100%. Single-piece production lacks "process stability", and each step is a new attempt. Any tiny variable (such as the batch difference of steel materials or the operating habits of workers) may cause the result to deviate from the target. This is why "single-piece production is the most difficult process to control in the world": you have to pay a huge cost for "uniqueness" but can't get "certainty" in return.
The essence of daily controllable events: Debugging and optimization of repeatable processes
Now let's look at our daily routines that we can "control", such as the time we get home every day. The underlying logic of this "controllability" is precisely "repeatable process": you can try different commuting routes (elevated roads vs. surface roads), adjust your departure time (10 minutes earlier vs. 10 minutes later), and change your means of transportation (driving vs. electric bike). Each adjustment is a "debugging". And because the process is repeatable, you can collect data (for example, leaving 10 minutes earlier can reduce traffic congestion by 20 minutes), optimize parameters (always take the elevated road and leave 10 minutes earlier), and finally stabilize your home - arrival time within a ±15 - minute range. This is exactly the same as the logic of mass - producing cars: only a repeatable process can achieve stability through debugging. The reason you can "control" your home - arrival time is not because you are "amazing", but because "commuting" is a process that allows repeated trial - and - error and optimization.
The "single attribute" of choosing a marriage partner: An unadjustable gamble
However, "choosing a marriage partner" exactly hits all the pain points of "single-piece production". First of all, the relationship is "non-repeatable" — you can't try living together with A for three years, then switch to B for another three years if it doesn't work out, and then switch to C for three more years (even if the law allows it, the emotional cost is unbearable). Secondly, the partner is an "individual with independent consciousness" — you can't, like adjusting automobile parts, say "change your personality to be gentler" or "adjust your living habits to go to bed early and get up early" and expect it to really happen. Finally, the cost of trial and error is so high that it's irreversible — a failed relationship consumes time, emotions, and confidence in intimate relationships, and these costs can't be "spread out" or "made up for".
The so - called "taking a gamble" actually means gambling on "compatibility": The person you choose just has enough "natural match" with you. For example, you share the same consumption outlook, your living habits are complementary, and your values are in line. You can live together without a lot of adjustment; or you're gambling on the "ability to adjust together": You're willing to face the differences together. Just like polishing a piece of uncarved jade, you slowly turn the "single - piece" relationship into a "stable system". However, there is no standard process for this kind of "adjustment", no parameter for reference, and even no "chance to make mistakes". In essence, this is a bet "with your future": You're betting that the person you choose is the one who "can go on with you without too much adjustment".
Conclusion: The boundary between the controllable and the uncontrollable
In fact, whether in manufacturing or in daily life, the core of "controllability" boils down to one thing: only a repeatable process can achieve stability through debugging. For a non - repeatable "one - off item", one can only rely on "gambles" or "luck". The time to get home can be controlled because commuting is a "repeatable process"; a car can operate stably because mass production is a "repeatable process"; however, choosing a marriage partner is uncontrollable because each relationship is a "unique one - off item" — you can't use the logic of "debugging" to solve the problem of "uniqueness". You can only gamble on the "compatibility" and the "ability to grow together". This is what makes life most interesting: the most precious things often don't make it onto the "controllable" list.