故事 · 一个专利审查员的反叛
Origin Story · A patent clerk's rebellion
1946 年,苏联海军的年轻专利审查员阿奇舒勒每天经手大量发明申请,渐渐生出一个大胆的念头:
创新会不会有可教、可学的规律?他不接受「天才灵感」这种说法,开始系统拆解专利。
几十年里,他和学生分析了超过 20 万份专利,发现真正有突破性的发明只占少数,
而它们用的「招数」翻来覆去就那 40 种 —— 分割、抽取、套娃、预先作用、变害为利……
他还发现:几乎每个技术难题都是一对矛盾(要轻又要强、要快又要稳),
于是把 39 个工程参数两两交叉,建成一张矛盾矩阵,告诉你「这类矛盾,前人常用哪几条原理破」。
这就是 TRIZ —— 让你站在二十万发明家的肩膀上,把碰运气变成查字典。
In 1946 a young Soviet Navy patent clerk named Altshuller was reading through stacks of invention disclosures every day, and slowly a heretical thought took shape:
could invention have teachable, learnable patterns? He refused to bow to the cult of genius and started taking patents apart, methodically.
Over the following decades he and his students worked through more than 200,000 patents, and the verdict was striking — only a small fraction were genuinely breakthrough,
and the moves those breakthroughs used kept boiling down to the same 40 maneuvers: segmentation, extraction, nesting, prior action, turning harm into benefit…
He also noticed that nearly every hard problem could be framed as a pair of contradictions (lighter yet stronger, faster yet steadier),
so he crossed 39 engineering parameters against each other and built the contradiction matrix — a table that tells you which principles past inventors have used to crack each kind of conflict.
That is TRIZ — climbing on the shoulders of two hundred thousand inventors, turning gambling into looking things up.
1 40 发明原理网格:点格子看案例 The 40 inventive principles: click a cell for an example
查矩阵或点原理Look up or click a principle2 TRIZ 的三块基石 The three pillars of TRIZ
技术矛盾:改善 A 必然恶化 B。把模糊难题翻译成一对可定位的工程参数。
Technical contradiction: improving A drags B down. Forces a vague problem into a locatable pair of engineering parameters.
矛盾矩阵:39×39 的查找表,行=改善、列=恶化,格内是前人验证的推荐原理编号。
Contradiction matrix: a 39×39 lookup table — rows are what to improve, columns are what gets worse, cells hold the principles past inventors trusted.
40 发明原理:分割、抽取、不对称、嵌套、预先作用…… 跨行业反复奏效的解题模板。
40 inventive principles: segmentation, extraction, asymmetry, nesting, prior action… cross-industry templates that keep paying off.
理想解(IFR):先想「最理想的结果」再倒推,避免一上来就被现有方案框死。
Ideal Final Result (IFR): name the perfect outcome first, then work backwards — so existing solutions don't cage your thinking from the start.
3 现实里的 40 原理 The 40 principles in the wild
原理 1 分割:模块化家具、可拆卸结构、组合工具 —— 把整体拆成可独立处理的部分。
Principle 1 · Segmentation: modular furniture, knock-down assemblies, combo tools — slice the whole into parts you can act on independently.
原理 13 反向:跑步机让人不动地动、自动扶梯把楼梯运起来 —— 把动的和静的对调。
Principle 13 · The other way around: treadmills let a runner stay still, escalators move the stairs — swap what moves with what stays put.
原理 7 嵌套:套娃、伸缩天线、嵌套量杯 —— 一物藏于另一物中,省空间。
Principle 7 · Nesting: Russian dolls, telescoping antennas, nesting measuring cups — tuck one object inside another to save space.
原理 10 预先作用:邮票背胶、预拉伸保鲜膜、安全气囊预充 —— 提前做好待命。
Principle 10 · Prior action: pre-glued stamps, pre-stretched cling film, pre-armed airbags — do the work ahead of time and stand ready.
一句话In one line
TRIZ 最反直觉的洞见是:创新不是灵感,而是可检索的经验。
你眼前那个「又要更强、又要更轻」的难题,在历史上早被无数发明家以各种形态解决过 ——
阿奇舒勒做的,就是把这些解法抽象成 40 条原理,再用矛盾矩阵索引起来。
于是你不再死磕「妥协一下吧」,而是被逼着问:能不能两个都要?前人是怎么两个都要的?
矩阵给你的不是答案,而是几把别人验证过的钥匙 —— 真正的发明,仍要你把抽象原理落到自己的产品上。
但它把创新从「等灵感降临」变成「系统化地翻字典」,这就是从碰运气到工程化的飞跃。
The most counter-intuitive lesson of TRIZ: invention isn't a flash of insight — it's searchable experience.
That "stronger and lighter at the same time" problem in front of you? It has already been cracked in countless forms by countless engineers —
what Altshuller did was abstract those moves into 40 principles and index them in a contradiction matrix.
So instead of muttering "let's compromise", you're forced to ask: can we have both — and how did past inventors have both?
The matrix doesn't hand you the answer; it hands you a few keys that have already worked for someone else — turning the abstract principle into your product is still your job.
But it converts invention from "wait for the muse" into "systematically thumb through the dictionary" — that is the leap from gambling to engineering.
常见误用Common mistakes
把矩阵当包治百病的答案机。它给的是方向(几条原理),落地还要你结合具体技术二次创造。
Treating the matrix as a magic answer machine. It only points the way — translating a principle into your specific technology still demands real engineering.
矛盾没定清就乱查。先把难题精确翻译成「改善什么、恶化什么」,矩阵才查得准。
Looking up the matrix before the contradiction is sharp. Translate the problem precisely into "what to improve" and "what gets worse" first — only then does the lookup hit the right cell.
只背 40 原理名词不练应用。原理要靠跨行业类比迁移,多看案例、多动手才用得活。
Memorizing the names of all 40 principles without ever applying them. Each principle has to be carried across industries by analogy — read cases, prototype, and only then will it stay alive in your hands.