A machine specification states positioning accuracy, spindle speed and travel. It does not state whether the shop can hold a flatness requirement on a large thin plate, keep a bore in tolerance after heat treatment, or produce the same result next quarter. Those outcomes depend on knowledge that is rarely written down.
This article sets out what that knowledge consists of and how sourcing and supplier-quality teams can evaluate it in a structured way.
What machine specifications do and do not tell you
Specifications describe capability under controlled conditions: a new machine, a test piece, ideal temperature. Production introduces material variation, fixturing, tool wear, thermal drift, multiple setups and operator decisions. The gap between rated accuracy and achieved tolerance on a real part can be large.
- Rated accuracy versus achieved tolerance on your part
- Machine age, maintenance and calibration state
- Environmental control in the shop
- Measurement capability relative to the tolerance
Tacit knowledge, process engineering and tooling
Tacit knowledge includes how an alloy batch behaves, when to change a tool before it degrades the surface, how to compensate for thermal growth during a long cycle, or which clamping sequence avoids distortion. It is usually held by a few experienced people and passed on through practice.
This knowledge is real but difficult to verify by questionnaire. It shows up indirectly in consistent results, thoughtful answers to process questions and low rework rates.
Process engineering converts a drawing into a sequence of operations, fixtures, tools and inspection steps. Tooling expertise covers cutting tools, molds, dies and fixtures. Together they determine yield and cost more than any single machine. Shops that design and maintain their own tooling often have tighter control over results, as discussed in our article on why Taiwanese machine shops modify their own equipment.
Tooling knowledge often decides cost as well as quality. A shop that has optimized cutting tools and fixtures for a part family can quote shorter cycle times and lower scrap, which shows up as competitive pricing that does not depend on low labor cost.
Process capability and production performance
The most objective way to evaluate a supplier's knowledge is through process capability data on critical characteristics. Cp and Cpk values, or simpler run charts from repeat production, show whether a process is centered and stable. For new parts, a first article inspection followed by a short capability run gives early evidence.
Where suppliers have no statistical records, a buyer can request measurement results across several pieces and lots of a comparable part.
Delivery history, nonconformance rates, corrective action records and customer longevity all indicate whether knowledge translates into reliable supply. A shop that has supplied the same part family to a demanding customer for many years has demonstrated something specifications cannot.
- On-time delivery and lead-time stability
- Nonconformance and return history
- Quality of corrective action reports
- Duration of key customer relationships
A structured technical evaluation
A practical evaluation combines document review, technical interview and part evidence. The interview should focus on a representative feature from the buyer's drawing and ask how the supplier would plan, fixture, machine, inspect and control it. The answers reveal depth quickly.
- Review a process plan for a representative part
- Walk through fixtures and tooling for comparable work
- Examine measurement methods and gauge capability
- Request first article and short-run capability data
Documenting manufacturing knowledge
For the buyer, documenting a qualified supplier's process through control plans, process flow diagrams and agreed change notifications protects the knowledge that made the supplier suitable. For the supplier, documenting capability in structured, comparable form makes it visible to buyers who would otherwise not find it. Some Taiwanese manufacturers now build manufacturing capability profiles for this purpose.
An evaluation example
A buyer is choosing between two shops for a ground and hardened shaft with tight runout and a specified surface finish. Shop A has a newer cylindrical grinder with impressive specifications. Shop B has an older machine but has supplied similar shafts to an automation customer for many years.
During the technical review, Shop B explains how it controls distortion through heat treatment sequencing, how it dresses the wheel for the finish, how often it checks runout during a batch, and provides measurement records from recent lots. Shop A confirms its machine accuracy but has limited records for comparable parts.
Neither is automatically the right choice. Shop B has stronger evidence; Shop A may become capable with process development. The point is that the decision rests on evidence about the process, and that evidence is not in the machine specifications.
Documentation also matters for continuity. If a key technician leaves or a machine is replaced, documented process parameters and control plans let the supplier restore performance faster and give the buyer a basis for verifying that nothing important changed.
Scoring tacit knowledge without guesswork
Teams often struggle to score qualitative findings. A simple scale for each critical characteristic helps: no evidence, verbal description only, process plan available, comparable-part data available, capability data on the actual part. Applied consistently, this turns a subjective impression into a comparable assessment across suppliers and regions.
- Level 0: no evidence
- Level 1: verbal explanation
- Level 2: written process plan
- Level 3: comparable-part measurement data
- Level 4: capability data on the actual part
Conclusion
Machine specifications set limits; knowledge determines performance within them. A capability assessment that combines process questions, fixtures, measurement and production data gives a much better prediction of supply performance than an equipment list. See also our article on why equipment lists don't tell you what a factory can actually make.
The takeaway
Evaluate suppliers on how they control your critical features and on production data, not on rated machine accuracy.