Designed & Made in the USA Since 1983
How wireless in-process data collection closes the gap between a compliance record and an early warning system.
When aerospace suppliers begin preparing for AS9100D registration, NADCAP accreditation, or their first First Article Inspection (FAI) per AS9102B, the conversation usually starts with processes, procedures, and training plans. What gets addressed late, sometimes too late, is the question of how measurement data actually gets from a gage into a quality system, and where in the production flow that measurement happens.
Both questions matter. A supplier that collects clean, traceable data but only at the end of the line, in a quality lab after the part is complete, has solved a traceability problem while leaving the detection problem. By then, the defect is already made, scrap is already cut, and the rework is already scheduled.
The suppliers best positioned for AS9100D surveillance, NADCAP audit, and customer source inspection are those collecting measurement data at every operation during production, before nonconformances compound. That shift from lab-based to in-process inspection has implications for how data collection infrastructure is designed from the start.
AS9100D Clause 7.1.5 establishes requirements for monitoring and measurement resources. Specifically, it requires that measuring equipment be calibrated, identified, protected from damage, and critically, that the results of measurement be valid and fit for purpose. Clause 9.1 further requires that the organization “evaluate and analyze data and information” to demonstrate conformance of products.
NADCAP auditors interpret these requirements with considerable rigor. Accreditation checklists for processes like heat treatment, NDT, and chemical processing include explicit scrutiny of how measurement data is recorded, whether it can be traced back to a calibrated instrument, and whether the data collection process itself introduces risk of error or tampering.
AS9102B, the First Article Inspection standard, requires that dimensional, functional, and material test results be documented against the design requirement for every characteristic at first article. The data package submitted to the customer must be complete and traceable. A single gap in the data record can hold up a program.
In all three frameworks, the implicit question is the same: How do you know the number in your quality system is the number the gage actually measured?
Consider a common scenario: a supplier running a first article on a machined aerospace component. The inspector uses a calibrated digital micrometer to measure wall thickness at twelve locations. The gage reads down to 0.0001″. The inspector reads the display, writes the value on a paper traveler, and a data entry clerk later keys it into the quality system.
That process introduces at least three failure modes:
Transcription error. Human re-entry of four-decimal-place values is error-prone. A single transposed digit changes a conforming result to a nonconformance, or worse, passes a nonconforming result as conforming. The AIAG MSA Reference Manual identifies measurement system errors as a significant source of process variation, and manual transcription is among the most controllable. For a detailed breakdown of what this costs in practice, see The True Cost of Manual Measurement Data Entry.
Traceability gap. When data is re-keyed, the link between the recorded value and the specific calibrated instrument that produced it is severed unless the inspector manually records the gage serial number and calibration status. In practice, that step is frequently skipped under production pressure.
Data latency. Paper-based collection at the end of a production run delays visibility until the damage is done. A dimensional drift that develops mid-run may not surface until end-of-batch lab inspection, at which point every part produced after the drift onset is potentially nonconforming. On a 500-part run at $38 per part, a tool or spindle issue developing at part 100 that goes undetected until final inspection exposes 401 parts to scrap or rework disposition, over $15,000 in material alone. Profitability is erased all before reinspection labor, CAPA, and customer expedite costs are factored in. Alternatively, an in-situ recording process could spot the anomaly at part 103. The same issue that scrapped $15,000 in materials affects only four parts. The process is the same. The detection point is the only variable.
None of these failure modes are exotic. They appear routinely in audit findings and CAPA records across the aerospace supply chain.
The more consequential shift isn’t just eliminating transcription error. It’s moving measurement earlier in the production flow. Quality lab inspection, however rigorous, is reactive in nature while In-process inspection at the machine or assembly station is proactive. The distinction matters enormously for cost of quality, and it’s increasingly what AS9100D auditors and aerospace customers expect to see in a mature quality system.
Relying on a quality depart to catch any and all errors at the end of an operation creates an intrinsic culture of defect acceptability. It implies that defects are acceptable to produce as long as they’re caught before shipment, a fundamentally reactive posture that embeds rework and scrap into the cost structure by design. Suppliers serious about process control don’t route conformance decisions through a downstream lab; they build inspection into the production flow so that out-of-tolerance conditions are surfaced and corrected before the next part is made.
The enabling constraint for in-process inspection has historically been infrastructure. Wired connections to a stationary data collection terminal work in a lab; they’re impractical on a dynamic machine floor where the inspector moves between stations and fixtures are constantly changing. While cabled solutions creates trip hazards and maintenance burden, wireless gage data collection provides the mobility and flexibility to realize in-process data collection.
With a wireless system like MobileCollect, inspectors roam freely across the production floor with a mobile device, collecting data at multiple points along the process and feeding it directly into cloud-based quality software like Net-Inspect. When multiple inspectors are working simultaneously, the volume of data flowing into the system increases significantly, giving quality teams far more data to work with when evaluating process stability. A robust wireless data collection system allows machine operators the ability to consistently capture feature data of equal quality to their inspection counterparts. That combination of inputs, analyzed in real time against control limits, is what makes out-of-control conditions visible before they produce a batch of nonconforming parts.
The mechanism that makes this actionable is Statistical Process control, SPC. Each measurement plots in real time against upper and lower control limits on an Xbar-R or I-MR chart. The system evaluates not just whether the current part is within specification, but whether the process producing it is stable and predictable. A reading trending toward a control limit, even while still conforming, is an actionable signal. That’s the operational difference between inspection and process control: inspection tells you what you made; SPC tells you what you’re about to make. This is, in essence, the foundation of your Process Capability Index, or Cpk.
For suppliers on a path toward higher automation, this architecture also establishes the data infrastructure that lights-out inspection requires. When measurement data flows from gage to quality system in real time, without human transcription, the collection process is no longer dependent on an inspector manually entering values at a terminal. That’s a meaningful step toward automated conformance decisions at the cell level.
A realistic example: a Tier 2 aerospace machining shop preparing for AS9102B FAI submission on a structural bracket. Twelve critical dimensions require inspection across ten parts. Rather than pulling finished parts to the quality lab, the inspection plan calls for dimensional checks at the machining cell, catching any drift before it propagates across the full lot.
The inspector uses calibrated digital calipers and a bore gage, each paired with a MobileCollect Mini Mobile Module EVO M3E. Moving freely around the workpiece while it is setup in the machine, the inspector measures each feature and presses the transmitter’s data button. The value transmits instantly to Net-Inspect, where it is logged against the correct characteristic, the correct part serial number, and the calibrated instrument ID, from the floor, in real time.
When a wall thickness reading on part three approaches the lower control limit, the system flags it before part four is staged. The operator adjusts the tool offset. The remaining seven parts run clean. Without in-process wireless collection, that drift would have been discovered in the quality lab after all ten parts were complete, with a potential scrap or rework disposition on the entire lot.
The FAI data package is assembled in Net-Inspect as inspection proceeds, not reconstructed afterward from paper travelers. When the customer’s designated engineering representative reviews the submission, every value is traceable to a specific calibrated instrument on a specific date, with no manual re-entry in the chain.
For NADCAP preparation, the same architecture supports audit readiness: the auditor asking “show me how this measurement was recorded and trace it back to the calibrated instrument” gets a complete, documented answer, along with evidence that measurements were taken at the point of production, not reconstructed after the fact.
The cost argument becomes sharper in ongoing production. Consider a 500-part production run over two days, a common scenario for Tier 2 machining operations. A tool wear or spindle issue developing at part 100 is invisible to a lab-only inspection process until the batch is complete. By then, 401 parts are potentially nonconforming. With wireless in-process collection and SPC monitoring at the cell, the same drift event triggers a control limit alert at part 103. Four parts are at risk. The operator corrects the offset. The remaining 397 parts run clean. Same process. Same drift. The detection point and the cost are the only things that differ.
The practical barrier to wireless in-process collection is lower than most suppliers expect, but there are legitimate decisions to work through:
Gage compatibility. Most digital precision gages produced in the last 15 years have Digimatic or RS-232 output capability, and a data output port is all that is needed to bring a gage online as a wireless data collection node. There are many brands and connectors available for digital gages and MicroRidge’s MobileCollect Selection Tool is a great tool to identify what wireless options exsist for your gages. Enter your gage make and model to confirm interface compatibility before deployment.Â
RF environment. Machine floors filled with electromechanical equipment, CNC enclosures, or other signal generator require a robust wireless solution. Industrial-grade wireless gage systems like MobileCollect use MicroRidge’s RM2.4 industrial wireless protocol, purpose-built for manufacturing environments where protocols such as Bluetooth and ATN fall short (see Why Bluetooth Fails in Manufacturing).
Inspection plan integration. Wireless collection delivers its full value when the inspection plan specifies which characteristics are measured in-process vs. at final inspection, with control limits defined for each. Without that structure, real-time data transmission doesn’t automatically produce real-time response. The operator needs to know what action to take when a value trends toward a limit.
Software integration. The collection system must write to the quality platform in a format the platform can receive. Net-Inspect among other SPC software providers support direct data import from wireless collection hardware, making integration with floor-level collection systems straightforward and eliminating any intermediate data handling step.
Calibration record linkage. The wireless module should capture the gage identifier at each measurement so the traceability chain, from the recorded value back to the calibrated instrument, is intact without manual entry. This is a configuration decision, not a hardware constraint, but it needs to be deliberate from the start.
“By the time a nonconformance is caught in the quality lab, the cost is already locked in, and frankly, if you’re relying on a downstream lab as your primary conformance gate, you’ve already accepted a broken system. The suppliers who are winning on quality are the ones who eliminated that gate entirely and moved inspection to the point of production. When you’re measuring in-line, at the machine, you catch the error before it becomes a part. That’s not just better compliance, it’s a fundamentally different cost structure. And once your data is flowing in real time, directly from the gage into your quality platform, you’re not far from the floor running itself.”
— Mike Dunlop, Founder, Net-Inspect
Net-Inspect is a cloud-based quality management platform purpose-built for the aerospace supply chain, supporting First Article Inspection, PPAP, and supplier quality workflows across AS9100D and NADCAP requirements. Net-Inspect and MicroRidge are integration partners.
AS9100D, NADCAP, and AS9102B all ultimately ask the same question in different forms: can you demonstrate that your measurement results are valid? Wireless in-process data collection answers that question more completely than lab-based collection does, not just because it eliminates transcription error, but because it closes the detection gap that end-of-line inspection leaves open.
For suppliers building a quality infrastructure that will hold up under registrar surveillance and customer source inspection, the investment in wireless collection and integrated quality platforms is a process control decision, not just a technology decision. It belongs in the same conversation as gage R&R studies, control plan development, and inspection plan design, early, and with the same rigor.
Suppliers who make that investment now are also positioning themselves for the next step, and it’s closer than most realize. When measurement data flows from gage to quality system in real time, inspection is no longer dependent on a human manually reviewing lab results and making a disposition decision. The conformance decision becomes a data event. Automated pass/fail logic at the cell level, operator alerts triggered by control limit exceedances, and quality records assembled without human intervention are all reachable with the infrastructure being deployed today. Lights-out manufacturing isn’t a dream state, it’s the endpoint of a system built upon a foundation automated datat collection.
Riley Tronson is President and owner of MicroRidge Systems, a role held since 2023. Riley brings a strong technical foundation to leadership in measurement solutions. An experienced entrepreneur, Riley has founded and grown multiple software companies, including a venture focused on developing iPhone applications, blending engineering expertise with innovative product development.