结论：检测越早介入，改动的成本越低

很多企业把检测安排在开发的最后阶段——产品定型了、模具开了、准备申报了，才送去检测。这时如果发现问题，改动的成本很高：可能要改模具、改结构、重新采购物料。

而如果在设计早期就用简易的方法验证关键假设，发现问题时改的只是图纸。

所以检测节点的安排原则是：把验证尽量前移，特别是对那些如果错了改动成本很高的设计决策。

各阶段的检测目的

阶段 检测目的 特点
概念阶段 验证关键假设 简易、快速、可用简化样件
设计阶段 验证设计方案 针对性强，用原理样机
样机阶段 全面验证 接近量产状态
定型阶段 型式检验 完整、正式、用于申报
试产阶段 工艺验证 验证量产一致性
量产阶段 例行检验 批次控制

各阶段的检测不是重复，而是目的不同。 早期的检测不追求全面和正式，追求的是快速回答关键问题。

概念阶段该验证什么

这个阶段验证的是那些「如果错了整个方案要推翻」的假设：

结构方案的承载可行性。 新的车架结构能否承受载荷。

材料选择的可行性。 选定的材料能否满足强度、重量、成本的组合。

关键机构的原理。 新的折叠机构、调节机构的原理是否可行。

尺寸与人体的配合。 关键尺寸是否合适。

验证方式可以很简化 ——用简单的样件、简化的加载、甚至计算加局部验证。目的是排除明显不可行的方案，不是得到精确数据。

设计阶段该验证什么

设计基本确定后，验证的是具体的设计参数：

关键部件的强度。 单个部件的静载和疲劳。

材料的实际性能。 供应商提供的材料是否达到规格。

配合与公差。 装配是否顺畅、间隙是否合适。

局部的功能。 制动、调节、折叠等机构的功能。

电气部分的基本特性。 电动产品的电机、控制器、电池的匹配。

这个阶段的验证针对性强，发现问题时改的是具体的尺寸、材料、或部件选型，成本仍然可控。

样机阶段该验证什么

样机接近量产状态，可以做较全面的验证：

整机的机械性能。 强度、疲劳、冲击、稳定性。

整机的电气性能。 电气安全、电磁兼容的摸底。

功能的完整验证。 各项功能在整机状态下的表现。

耐久试验。 时间较长，应当尽早启动。

环境适应性。 温度、湿度、防水。

用户试用。 实际使用中的问题。

「耐久试验尽早启动」是时间管理上的要点。 耐久试验周期长，如果等到定型才做，结果出来时已经来不及改。建议在样机阶段就启动，与其他工作并行。

定型阶段的型式检验

这是正式的、用于申报的检验：

覆盖全部适用标准。 按目标市场的标准清单。

样品是定型状态。 用定型工艺生产。

报告要完整。 含条件、数据、样品信息。

一次做全。 考虑各目标市场的需求，避免后续补测。

这个阶段发现问题的代价最大，所以前面几个阶段的验证做得充分，这一阶段的风险就低。

试产阶段的工艺验证

批次一致性。 试产批次内和批次间的差异。

工艺参数的确认。 各工序参数的适宜范围。

检验方法的确认。 出厂检验的项目和方法。

关键特性的识别。 哪些参数需要重点控制。

这个阶段的数据是量产质量控制的基础。

早期验证的价值

把验证前移的收益：

改动成本低。 图纸阶段改比模具阶段改便宜得多。

时间损失少。 早期发现问题不影响整体进度。

方案选择有依据。 多个方案时用数据选择而不是靠判断。

风险早暴露。 避免后期的意外。

减少返工。 定型检验一次通过的概率提高。

代价是： 早期验证需要投入，而且用的是不完整的样件，数据的代表性有限。

但相比后期发现问题的代价，这个投入通常是划算的。

与设计迭代的配合

快速反馈。 早期验证追求快，结果能及时反馈给设计。

针对性强。 每次验证针对当前的设计问题，不追求全面。

方法可简化。 不必都用正式的标准方法，能回答问题即可。

记录要保留。 早期验证的数据在后续的设计论证中有用。

与正式检验衔接。 早期数据可以作为正式检验的预期参考。

节点安排的示例思路

把一个电动轮椅的开发节点大致排一下：

概念阶段。 车架结构方案的承载验证、电池与电机的匹配核算。

设计阶段。 关键部件强度、材料性能确认、控制器与电机的实测匹配。

样机阶段。 整机强度与稳定性、EMC 预扫、电气安全摸底、耐久试验启动、用户试用。

定型阶段。 完整的型式检验，覆盖目标市场的标准清单。

试产阶段。 批次一致性、出厂检验方法确认。

量产阶段。 例行检验、定期型式复检。

这个思路可以按产品的具体情况调整，原则是把代价大的验证前移。

与申报节奏的配合

检测节点还要与申报的节奏配合：目标市场的申报要求；报告的时效要求；申报所需的资料清单；以及审查中可能的补充要求。

建议在定型检验前确认目标市场的完整标准清单，一次做全，避免申报过程中发现缺项再补测。

我们的做法

我们承接的委托中，多数是定型阶段的正式检验，但早期阶段的摸底测试同样可以做 ——简化的加载试验、材料性能验证、局部功能测试，这类测试周期短、成本低，对开发的价值明显。

对于计划做正式型式检验的委托方，我们建议在样机阶段先做一轮摸底 ——特别是电气安全和电磁兼容，这两项的整改涉及电路和结构，越晚发现代价越高。

耐久试验我们建议尽早启动，因为它周期长，往往是开发进度的制约项。

如果你在规划新产品的开发节点，想理清检测怎么安排，可以把开发计划和产品方案发过来一起讨论，或者直接联系：132 4819 8029。检测能力见服务介绍，产品分类见手动轮椅检测与电动轮椅检测，标准信息见标准查询。

English version

Conclusion. Many manufacturers schedule testing at the end of development, once the product is settled, tooling is cut and submission approaches. Problems found then are expensive to fix: tooling may need modifying, structures redesigning and materials repurchasing. Verify key assumptions early with simple methods and problems found change only a drawing. The principle for scheduling test milestones is therefore to move verification as early as possible, particularly for design decisions that would be expensive to reverse.

Purpose by stage. At concept stage, verification tests key assumptions, using simple and rapid methods on simplified specimens. At design stage, it verifies the design approach, being targeted and using functional prototypes. At prototype stage, it verifies comprehensively on units close to production. At design freeze, type testing is complete, formal and used for submission. At pilot production, it verifies the process and production consistency. And in production, routine inspection controls batches. These are not repetition; their purposes differ. Early testing does not seek completeness or formality but quick answers to key questions.

What to verify at concept stage. Verify the assumptions whose failure would overturn the whole approach. Load feasibility of the structural concept, meaning whether a new frame arrangement can carry the load. Material feasibility, meaning whether the chosen material meets the combination of strength, weight and cost. The principle of key mechanisms, meaning whether a new folding or adjustment mechanism can work. And dimensional fit with the body, meaning whether key dimensions suit. Verification can be very simplified, using rough specimens, simplified loading or calculation with local verification. The purpose is eliminating clearly unworkable approaches rather than obtaining precise data.

What to verify at design stage. With the design largely settled, verify specific parameters. Component strength, covering static and fatigue loading of individual parts. Actual material performance, confirming that supplied material meets specification. Fits and tolerances, confirming assembly is smooth and clearances appropriate. Local function, covering braking, adjustment and folding mechanisms. And basic electrical characteristics for powered products, covering the match between motor, controller and battery. Verification here is targeted, and problems found change a dimension, a material or a component selection, keeping cost controllable.

What to verify at prototype stage. Prototypes approach production condition and allow fuller verification. Whole-product mechanical performance, covering strength, fatigue, impact and stability. Whole-product electrical performance, surveying electrical safety and electromagnetic compatibility. Complete functional verification in assembled condition. Durability testing, which takes time and should start early. Environmental suitability, covering temperature, humidity and water ingress. And user trials, revealing problems in real use. Starting durability testing early is the key scheduling point: durability takes a long time, and leaving it until design freeze means results arrive too late to act on. Start it at prototype stage, in parallel with other work.

Type testing at design freeze. This is the formal testing used for submission. Cover all applicable standards for the target markets. Use samples in definitive condition, made by the definitive process. Make reports complete, with conditions, data and sample information. And do it all in one pass, considering every target market to avoid supplementary testing. Problems found at this stage cost the most, so thorough verification at earlier stages keeps the risk here low.

Process verification at pilot production. Batch consistency covers variation within and between pilot batches. Process parameter confirmation establishes the appropriate range for each operation. Inspection method confirmation establishes the items and methods for final inspection. And key characteristic identification establishes which parameters need close control. Data from this stage underpin quality control in production.

The value of early verification. Changes cost less, since changing a drawing is far cheaper than changing tooling. Time lost is less, since problems found early do not affect the overall schedule. Option selection has a basis, choosing between approaches on data rather than judgement. Risks surface early, avoiding late surprises. And rework reduces, raising the probability of passing type testing first time. The cost is that early verification requires investment and uses incomplete specimens whose data have limited representativeness. Against the cost of finding problems late, however, that investment is usually worthwhile.

Fitting design iteration. Feedback must be quick, so that early verification results reach the designers in time. Verification must be targeted at the current design question rather than comprehensive. Methods can be simplified, not necessarily following formal standard methods as long as they answer the question. Records should be kept, since early data support later design justification. And it should connect to formal testing, with early data serving as expected values.

How we handle it. Most of our work is formal testing at design freeze, though we can equally perform early survey testing: simplified load tests, material performance verification and local functional testing, all quick and inexpensive and of obvious value during development. For clients planning formal type testing we suggest a survey round at prototype stage, particularly on electrical safety and electromagnetic compatibility, since remediation there involves circuits and structure and gets more expensive the later it is found. We suggest starting durability testing early, since its duration often constrains the development schedule.

Send us your development plan and product concept and we will work through the test milestones. Phone or WeChat: +86 132 4819 8029.