起源故事 · 别再让三套工具各喊各的
Origin Story · Stop Letting Three Toolsets Pull in Three Directions
上世纪末到本世纪初,企业里常同时养着三拨改善人马:搞精益的追着浪费砍,搞六西格玛的盯着缺陷降,
搞 TOC 的找瓶颈。问题是,精益和六西格玛都很能干,却常常在不影响整体产出的地方拼命用力——
把一道本就不缺活的工序优化到飞起,系统总产出却纹丝不动。Reza Pirasteh 等人提出 TLS(TOC–Lean–Six Sigma),
做法极简:先用 TOC 找到约束,再把精益和六西格玛全压到约束上。实证研究里,TLS 带来的财务收益,
数倍于单独跑精益或六西格玛——不是工具更强了,而是火力对准了真正决定产出的那一点。
From the late 90s into the 2000s, enterprises often kept three separate improvement crews on the payroll:
Lean teams chased waste, Six Sigma teams chased defects, and TOC teams chased bottlenecks.
The trouble was that Lean and Six Sigma, both genuinely powerful, kept pouring effort into places that didn't move the overall throughput —
optimizing already idle steps to perfection while the system output didn't budge. Reza Pirasteh and colleagues proposed TLS (TOC–Lean–Six Sigma) with a deceptively simple recipe:
use TOC to find the constraint first, then drop the full weight of Lean and Six Sigma directly on it. Empirical studies showed TLS delivering financial returns several times larger than running Lean or Six Sigma alone —
not because the tools were stronger, but because the fire was finally aimed at the one point that actually determined throughput.
1 五道工序产线:火力撒在哪 A Five-Step Line: Where the Fire Lands
到处撒网Spread Thin2 同样的投入,产出与 ROI 差几倍 Same Spend, Multiples of Throughput and ROI Apart
撒网把资源喂给了非约束工序——它们更快了,但系统产出由约束决定,整体几乎没动。TLS 把同样的资源压在约束上,每一分投入都直接抬升瓶颈,产出与 ROI 一起拉开差距。 Spreading the budget feeds non-constraint steps — they get faster, but system throughput is fixed by the constraint, so the overall number barely moves. TLS aims the same budget at the constraint, so every dollar lifts the bottleneck directly — throughput and ROI pull away in lockstep.
3 现实里的 TLS TLS in the Wild
改善体系整合:与其让精益、六西格玛各干各的,不如让 TOC 统一指挥,三套工具服务同一目标。
Unified improvement program: instead of letting Lean and Six Sigma run as separate kingdoms, put TOC in charge of targeting and have all three toolsets serve a single objective.
聚焦式改善:项目立项先问「这改的是约束吗」,不是约束就排后,避免在无关工序上烧钱。
Focused project selection: every charter starts with "is this on the constraint?" If not, it goes to the backlog — no more burning budget on irrelevant steps.
避免遍地开花:黑带项目最大的浪费,是优化了一堆「局部最优」却没动整体——TOC 先帮你筛。
No more shotgun improvement: the biggest waste in a Black Belt program is a portfolio of local optima that never moves the global number — TOC pre-filters them out.
最大化改善 ROI:约束处每提升 1%,全系统产出就提升 1%;非约束处提升 1%,往往等于 0。
Maximum improvement ROI: a 1% gain at the constraint lifts the whole system by 1%; a 1% gain anywhere else usually lifts the system by 0.
一句话In One Line
精益和六西格玛都是强大的工具,但工具再强,用错地方就是零回报——
把非约束工序优化到极致,系统产出依旧被瓶颈卡住。TLS 的全部精髓,是给这两套工具装上一台
TOC 瞄准镜:先定位约束,再开火。于是「1(定位)＋1(提速)＋1(稳定)」不再是简单相加,
而是乘法效应——同样的人、同样的工具、同样的预算,回报却能翻几倍。
改善的难点从来不是「能不能改」,而是改在哪里才算数。
Lean and Six Sigma are both powerful toolsets, but no toolset, however sharp, survives being aimed at the wrong target — the return is zero.
Polish a non-constraint step to perfection and the system output is still pinned by the bottleneck. The entire point of TLS is to bolt a
TOC scope onto both toolsets: locate the constraint first, then open fire. With the scope in place, "1 (focus) + 1 (speed) + 1 (stability)" stops being simple addition and becomes a
multiplicative effect — same people, same tools, same budget, several times the return.
The hard part of improvement was never "can we improve it" — it's where the improvement actually counts.
常见误用Common Mistakes
把 TLS 当成「三套工具都做一遍」。核心是「先 TOC 聚焦,再用 Lean/Six Sigma 攻约束」,不是堆工具。
Treating TLS as "run all three toolsets once". The core is "TOC focuses first, then Lean and Six Sigma attack the constraint" — not a tool pile-up.
哪个项目好看做哪个,不看是否约束。立项先问「这改的是约束吗」,非约束的改善多半不计入产出。
Picking projects by what looks good, not by whether they target the constraint. Every charter starts with "is this on the constraint?" Non-constraint gains rarely show up in throughput.
约束突破后还在原地继续优化。约束会转移,要重新用 TOC 定位下一个,再调火力。
Keeping the fire on the old constraint long after it's been elevated. The constraint moves — relocate it with TOC, then re-aim the fire.