Heat Treating Operations Software: How Do You Answer a Pyrometry Auditor in Ten Minutes Instead of a Week?

The audit question that takes a week to answer

An auditor picks a job from your shipping records at random. Parts run in March, carburised in furnace 6. He asks for four things, and he asks for them together.

The load record with the chart showing that the load reached temperature and held it for the required soak. The system accuracy test that was current for that instrument on that date. The temperature uniformity survey that qualified furnace 6 for the operating temperature that job ran at. And the usage history of the specific thermocouple used in that test, showing it was within its permitted use limits.

In most shops those four things live in four places. Load charts are in a filing cabinet by furnace and month, or on a recorder's memory card. Test results are on a form in a binder in the quality office. Calibration certificates are PDFs in an email folder, filed by the vendor's invoice number. Thermocouple usage is a tally on a clipboard by the test bench, if it is anywhere. Assembling one traceable chain across all four takes the quality manager most of a day, and the auditor is going to pick five more jobs.

The exposure is not the finding. It is what a finding costs. Aerospace approvals are the reason your most profitable work walks in the door, and a pyrometry non conformance can put a furnace out of service until the corrective action clears. Losing a process approval does not cost you an audit fee, it closes you out of a customer base.

Why this operation breaks on general shop software

A general job shop system models an order, a routing and an operation. Heat treating adds three things it has no place for.

The first is furnace qualification as a live state. A furnace is not simply available. It is qualified for a temperature range, under a class and instrumentation type, until a specific date, and its qualification lapses on a schedule that depends on those attributes and on its recent survey history. Scheduling a job into a furnace whose survey expires the day before the job runs is a mistake a computer should catch and a whiteboard never will.

The second is the sensor. Thermocouples are consumables with rules attached: type, permitted uses or hours at temperature, calibration where applicable, and a correction factor that applies to readings taken with them. A general system has an inventory item. It does not have a sensor whose remaining permitted uses gate whether a test performed with it is valid.

The third is the chart. The proof that a load did what the recipe said is a recorder trace, and it must attach to the job, the load, the furnace and the parts inside it, not to a folder named by date. Every shop that files charts by furnace and month is one customer question away from a manual search.

What Super Systems and Datapaq actually leave you doing

Super Systems Inc is the deepest name in this space and its equipment and software genuinely handle the pyrometry side: data acquisition from furnace instruments, survey support and compliant recording. If your problem is capturing and retaining accurate temperature data, that is what it is built for.

Fluke Process Instruments Datapaq is the standard answer for through process temperature profiling, particularly where a system travels with the load. Excellent at what it does, which is producing a trusted profile of what the product actually experienced.

Neither is a shop operating system, and that is the gap. They hold the temperature truth. They do not hold the job that walked in this morning, the customer's specific recipe and hardness requirement, the racking and load build, the quench delay limit, the outside processing that follows, the certificate you have to issue, or the invoice. In practice the shops we work with run one of these products for the pyrometry data and then hold everything else in a mix of paper travellers, a general accounting package and a scheduling whiteboard. The audit pain sits precisely in the joins between them.

What a custom build has to include

Start with the equipment register. Every furnace with its class, instrumentation type, qualified operating range, control and recording instruments, and the survey and test history that keeps it qualified. From that register the system computes the next due date for each required test rather than asking someone to remember an interval, and it warns before expiry rather than after. Qualification state then becomes a scheduling constraint: a job cannot be planned into a furnace that will not be qualified for its temperature when it runs.

Then the sensor register. Thermocouples by type and role, with permitted use limits, accumulated usage, correction factors and expiry. A test recorded against a sensor past its limit should be rejected at entry, because finding that out during an audit is the expensive version of the same discovery.

Then the load. Parts, quantity, customer, recipe, racking, the furnace it ran in, start and end times, operator, and the recorder trace attached automatically rather than filed. Include the small things that carry real risk: quench delay from furnace to quench, load thermocouple placement, and the hardness or metallurgical test results that follow. The load record is the object an auditor and a customer both ask about, so it should be the object everything else hangs from.

Then the traveller. The job moves through preheat, the treatment cycle, quench, temper, straightening, testing and often outside processing before it ships. Whatever your shop actually does needs to be visible in one place, because the second most common audit question after pyrometry is where a specific batch of parts was between two dates.

Then certificates generated from captured data, not typed. A certificate assembled by hand is where transcription errors enter your traceability chain, and it is the document your customer keeps.

One more piece is worth naming because shops consistently underestimate it: automated data capture from recorders and instruments of several different makes. Most shops run a mix accumulated over 20 years. Reading them into one system is unglamorous integration work, and it is also the difference between a system people use and a system people type into twice.

Cost, timeline and what moves them

The first release runs $50,000 to $110,000 and ships in 10 to 16 weeks. That is the equipment and sensor registers with automatic scheduling, test and survey record capture, load records with attached charts, and audit retrieval that answers the four part question above in one screen. Shops that hold aerospace approvals should build in this order specifically, because it removes the highest cost risk first.

The full shop system with quoting, order intake, scheduling across furnaces, customer recipes with revision control, outside processing, certificates, shipping and accounting integration runs $140,000 to $300,000 across 5 to 10 months. What pushes it up: multiple plants, a wide mix of recorder and controller makes to integrate, automotive assessment requirements alongside aerospace, and metal finishing lines with their own chemistry and tank controls. What holds it down: one plant, and accepting that the first release covers pyrometry compliance and load records before it touches quoting.

When you should not build this

A one or two furnace commercial shop doing commercial grade work with no aerospace or automotive approvals does not need custom software. Keep the binders, keep them tidy, and buy a recorder that stores digitally. The build case begins with approvals, with more furnaces than one person can track expiry dates for, or with the realisation that your quality manager spends more time assembling evidence than improving process.

How to choose a developer

Ask them to explain, without prompting, why a furnace has a qualified temperature range and what happens to scheduling when a survey expires. If they treat furnace availability as a simple calendar, they will build you a scheduling tool that books jobs into non compliant equipment.

Ask how a thermocouple's remaining permitted uses affects the validity of a test recorded with it. A developer who has worked in this space will answer immediately. One who has not will suggest a maintenance reminder, which is not the same thing.

Ask what recorder and controller hardware they have actually pulled data from. Ask for the make and the interface, whether that was a file drop, a serial protocol or a network read. This is the part of the project most likely to run long and it is the part vendors are vaguest about.

Ask who owns the code, the repository, the cloud accounts and the historical records, and get it in writing before kickoff. At Digital Heroes the client owns all of it from the first commit. Pyrometry records have retention obligations measured in years and your customers may ask for them long after any software contract ends, so they must live somewhere you control outright.