Emma: So imagine a massive pallet of packaged food product, right? It’s being loaded onto a delivery truck, like right now.
Ryan: The engine is running.
Emma: Exactly. The engine is running, the driver is ready to go. And honestly, the only thing standing between that manufacturer and a, you know, a multimillion dollar headline making pathogen recall.
Ryan: Oh.
Emma: Is a single grease stained handwritten paper checklist from last Tuesday.
Ryan: It’s a terrifying reality, but it is the reality of modern manufacturing.
Emma: It really is. I mean, if you want to understand how a business is actually running, you have to skip the fancy executive dashboard. You need to go find that overflowing, slightly dented metal filing cabinet.
Ryan: The one sitting right in the corner of the quality control office.
Emma: Yes. Just stuffed to the brim with smudged ink and coffee rings.
Ryan: Which is basically the universal monument to operational anxiety. Yeah, because every single one of those pieces of paper represents a really high stakes moment.
Emma: A moment where a human being had to physically verify that a product wasn’t going to, you know, fail or break or contaminate a whole supply chain.
Ryan: Exactly. And managing that mountain of paper is just a logistical nightmare. I mean, relying on it means relying on a profoundly fragile system.
Emma: And usually the only thing keeping that entire paper based house of cards from just completely collapsing is tribal knowledge.
Ryan: Oh, for sure.
Emma: Like it’s that one seasoned floor supervisor who just happens to know that the heavy duty mixer always runs a little hot on Tuesdays. Or, you know, that a specific vendor’s metal brackets always need a second glance because they have a history of microscopic fractures.
Ryan: Right. And none of that context is actually in a database anywhere. It is entirely locked inside someone’s head,
Emma: which is an incredible vulnerability. I mean, when that supervisor retires, or honestly even just takes a week off for vacation, all that critical quality control context just walks right out the door
Ryan: and the organization is left totally blind.
Emma: Which brings us perfectly to the mission for today’s deep dive. We are unpacking a stack of sources on an application called Quality Inspector.
Ryan: Right. Which is built by Insight Works for the Microsoft Dynamics 365 Business Central platform.
Emma: Exactly. And look, if you are dealing with the daily headache of quality control, whether you are, you know, assembling complex bicycles or testing food batches for bacteria, this is fundamentally about how you bridge that massive gap between physical reality and digital records.
Ryan: Yeah. So first we are going to explore how a team actually digitizes that messy
Emma: paper trail the paper problem.
Ryan: The paper problem, Exactly. And then we’ll dive into a specific chapter on their newest release, version 7.26, to answer one really critical question. What does it actually take to get from a failed test to a closed corrective action?
Emma: And where does that process usually just completely break down? Okay, let’s unpack this, because before we can even talk about, you know, advanced analytics and correlation, we have to talk about how data actually gets into the
Ryan: system in the first place without causing a total mutiny on the shop floor.
Emma: Right. Because moving from paper to electronic QC is notoriously difficult. Anyone who has tried to roll out new software in busy industrial environment knows this.
Ryan: It almost always sparks a massive battle between the IT department and the shop floor.
Emma: I mean, I can see why you can’t just hand a forklift driver or a welder fragile laptop and expect a standing ovation. You are essentially invading their workspace.
Ryan: You are introducing a massive amount of friction into their day. I mean, the workers on the floor are accustomed to clipboards for very logical reasons. A clipboard is, is tactile, it’s highly portable, it doesn’t crash, it doesn’t run out of battery, and most importantly, you know, you don’t have to take your heavy safety gloves off to use it.
Emma: That’s a huge point.
Ryan: It is when you force a worker to suddenly drop that clipboard and switch to navigating a keyboard and a, you know, a complex software interface.
Emma: Yeah.
Ryan: You’re asking them to learn a completely
Emma: new workflow when all they really want to do is inspect the physical part and just move on to the next task.
Ryan: Exactly. So the software essentially has to adapt to the worker, not the other way around.
Emma: And Quality Inspector tackles this by avoiding the whole cold turkey software transition entirely. Right.
Ryan: Yeah. They use Microsoft’s Azure Document Intelligence, which
Emma: is highly advanced optical character recognition, or ocr.
Ryan: Right. And the application of this technology here is really a masterclass in change management. It directly addresses the psychological hurdle of digital transformation.
Emma: Let’s visualize how this actually plays out on the floor. Think about a standard 5s workplace organization checklist. Just the basic daily form making sure the floor is swept and tools are put away.
Ryan: Or even a really messy handwritten inspection sheet.
Emma: Right. The sources actually give a great example of form filled out by Lisa in the paint area. It’s got check marks, scribbled notes in the margins, and literally probably some paint smudges on it.
Ryan: Absolutely.
Emma: So instead of that worker having to sit down at a terminal and manually type all that raw data in, they can simply Take a rugged drop proof tablet and snap a photo of that messy piece of paper. Paper.
Ryan: And the application scans the image, reads the handwritten text, and maps the physical check marks to digital yes or no toggles.
Emma: It just instantly generates the populated data entry form right inside Business Central.
Ryan: It digitizes the entire physical document in a matter of seconds. You bypass the manual data entry bottleneck completely.
Emma: It’s essentially like. It’s like teaching your smartphone to instantly understand your doctor’s notoriously messy handwriting.
Ryan: That’s a great analogy.
Emma: And then having it automatically update your digital medical chart based on those scribbles. You are basically giving your central database eyes to read the clipboard.
Ryan: And consider why this matters for team morale. You aren’t forcing your workforce to immediately abandon their deeply ingrained physical habits.
Emma: Right. The clipboards.
Ryan: Yeah. You are allowing them to operate exactly as they always have while the software builds this hidden digital shadow of their paper process in the background.
Emma: It bridges the gap until the workforce is comfortable enough to eventually just drop the paper altogether.
Ryan: Exactly. And use the digital interfaces natively.
Emma: Okay, so we’ve solved the data entry bottleneck. The handwritten scribbles from Lisa and the paint department are now properly structured rows and columns in the database. But, I mean, collecting data is just the prerequisite.
Ryan: Right. What does the engine actually do with those numbers once they’re inside?
Emma: Right. And this is where the system really has to prove its flexibility. Because manufacturing environments are wildly inconsistent, the
Ryan: application has to handle completely different paradigms of testing.
Emma: To put that in perspective, consider two very different inspection scenarios from the sources. On one hand, you have a simple
Ryan: bicycle assembly checklist which relies on basic Boolean logic.
Emma: Right. Simple yes or no questions. Are the wheels attached properly? Yes. Is the serial number tag securely fastened? Yes. It’s a very straightforward binary pass or fail hurdle before a product gets put in a box.
Ryan: But on the other end of the spectrum, you have highly sensitive operations like a pathogen test.
Emma: Oh, wow. Yeah.
Ryan: Imagine an inspection happening at a commercial mixer work center where a team is checking for coliform or E. Coli in a massive batch of organic food product.
Emma: That isn’t a simple yes or no.
Ryan: Not at all.
Emma: Yeah.
Ryan: That requires measuring specific numerical values, checking those values against acceptable chemical tolerances, and calculating, you know, complex statistical averages across multiple different samples taken from the same batch.
Emma: And sampling is where things get really intricate. There is a statistical concept built into the system called Aqilo. Right?
Ryan: Yes. Acceptable quality limits.
Emma: So the system looks at the total batch size and it automatically calculates a total default Sample size based on mathematical parameters. It tells the inspector exactly how many items they need to pull to get a statistically significant result.
Ryan: However, there is a built in function that allows the human inspector on the floor to manually override that algorithm.
Emma: Like in the source demo, where a 100 kilogram batch was supposed to be tested.
Ryan: Right. And the inspector manually changed the parameters to test exactly 10 samples with a strict limit of only one allowed failure.
Emma: Wait, hold on. Let me push back on this for a second.
Ryan: Okay?
Emma: Yeah, because I’m stuck on the logic here. If a company is investing heavily in sophisticated quality control software, isn’t the entire point to make the process fully automatic?
Ryan: You would think so.
Emma: I mean, to strip the human error and bias completely out of the equation, why let a floor worker override a statistically sound algorithm? Doesn’t that sort of defeat the purpose of having the software in the first place?
Ryan: What’s fascinating here is that it actually acknowledges the chaotic reality of the shop floor. I mean, real world manufacturing is never perfectly uniform.
Emma: True.
Ryan: Sometimes a new batch of raw materials from a supplier behaves slightly differently than the last one. Or sometimes ambient environmental factors like severe humidity in the warehouse alter the testing conditions.
Emma: So a rigid system actually creates more problems than it solves?
Ryan: Precisely. True operational software augments human expertise rather than blindly replacing it. If the system were just a rigid black box that refused to let the inspector adjust for localized context, the inspector would inevitably start working around the system.
Emma: Oh, I see. They would do the extra tests off the books.
Ryan: Exactly. Or they would manipulate the data to force the system to accept reality. By allowing the manual override, the software keeps the human in the loop, but forces them to document their adjustment.
Emma: It captures that vital human judgment as part of the official auditable digital record.
Ryan: It transforms tribal knowledge into documented knowledge
Emma: that makes total sense. And once those results, whether algorithmic or human adjusted, are actually finalized, the system takes over to handle the grading.
Ryan: Right. It automatically grades the specific lot of products. So it assigns it a grade A, B or C based entirely on the recorded test values.
Emma: And the crucial part is that those grades trigger immediate automated consequences throughout the ERP system.
Ryan: Think about the real world impact of this. Let’s say a massive batch of specialized materials goes through testing and receives an automated grade C. Okay. The parameters can be set up so that grade C material is physically and systematically blocked from being allocated to any external sales order.
Emma: So a warehouse worker physically cannot scan it onto a delivery truck headed for a customer.
Ryan: Right. But that same grade C material might still be perfectly acceptable for internal Consumption in a completely different, less critical manufacturing process down the hall.
Emma: Oh, wow. So the system handles that routing nuance
Ryan: instantly, without a manager needing to desperately run onto the floor or blast out a company wide email shouting, you know, do not ship blot402, which is huge for efficiency.
Emma: And when human communication is necessary, the system integrates seamlessly with Microsoft Teams.
Ryan: Yeah. If a critical test fails, the application can automatically push an adaptive card directly into a designated teams channel.
Emma: An adaptive card being a sort of interactive, structured summary widget right inside the chat interface.
Ryan: Exactly. This allows stakeholders who don’t spend their day logged into the core database, like a VP of Operations or a floor manager, to immediately see the critical details
Emma: of the failure right on their phone or desktop. They have visibility into the fire without needing a dedicated software license or deep technical training.
Ryan: Exactly.
Emma: So, okay, we’ve successfully digitized the messy paper trail. The data is securely and business central. The sampling is flexible, and the grading is automated.
Ryan: But just storing data doesn’t actually improve a product.
Emma: Right. The real question is, how does a team use all this new digital data to stop the next failure before it happens?
Ryan: Which brings us explicitly into what is brand new, the core problem with traditional quality control and the major paradigm shift introduced in the application’s newest release, version 7.26.
Emma: Because up until this update, quality control teams were largely operating in a reactionary mode. They were looking at failed tests one isolated incident at a time.
Ryan: A widget fails the stress test, you fix or scrap the widget. A batch is contaminated, you throw out the batch.
Emma: It’s like bailing water out of a leaky boat with a teacup without ever stopping to actually look for the hole in the hull. You spend all your time reacting to the water, but you never fix the source of the leak.
Ryan: That is the perfect analogy. And identifying a broader pattern, realizing that a specific leak is actually getting worse, usually happens entirely by accident.
Emma: How much time does a quality team typically spend just trying to figure out if a failure is a one off or an actual trend?
Ryan: A massive amount of time. Because right now, pattern finding happen by tribal knowledge, a floor manager might suddenly realize, wait a minute. Every time we use this specific vendor for our titanium screws, the final assemblies seem to fail.
Emma: But by the time a human being mentally connects those dots, the company has already hemorrhaged thousands of dollars in scrapped materials and wasted labor.
Ryan: Right. So to solve this, version 7.26 introduces a failure correlation analysis page, which is huge. It essentially hardwires that elusive pattern finding capability directly into the system itself. It scans across all the historical test data to spot the hidden correlations that a human would easily miss or might
Emma: take six months to finally notice.
Ryan: Exactly. It actively points out that, hey, failure rates are suddenly spiking. And more importantly, all those recent failures share a common denominator.
Emma: Like they all used this specific machine or were handled by this specific operator,
Ryan: or originated from this specific raw material lot. It moves the entire QC department from a defensive posture to an offensive one.
Emma: You aren’t just reacting to a failed test. You are tracking the underlying variables to eradicate the root cause.
Ryan: And speaking of underlying variables, a failure pattern often points back to the physical tools being used on the floor.
Emma: Oh, right. Which brings up a massive compliance guardrail. In this new release regarding tool calibration.
Ryan: In any regulated industry, the physical tools you use to measure quality must be rigorously calibrated on a set schedule. We are talking about digital scales, calipers, torque wrenches, temperature probes. Right. If an auditor discovers that a tool has fallen out of its calibration window every single test performed with that tool since its expiration date is suddenly invalidated,
Emma: that is an absolute nightmare scenario. You could be forced to recall months of shipped product just because someone forgot to certify a thermometer.
Ryan: It destroys trust and costs a fortune. So what the new update does is enforce tool calibration at the exact point of the test.
Emma: It physically blocks the use of any tool with an expired calibration date.
Ryan: The software will not even allow the inspector to record the test values if the system recognizes that the caliper they selected is past its certification window.
Emma: So a worker couldn’t falsify or accidentally corrupt the record even if they tried. It stops the error before it ever poisons the database. Okay, so we’ve spotted the failure patterns and we know our measurement tools are legally calibrated. But identifying the problem is really only half the battle.
Ryan: The other half is the administrative nightmare of filing the paperwork to actually authorize a fix.
Emma: Which brings us to the largest cluster of changes in the v7.26 update. The handling of non conformance reports, or NCRS.
Ryan: The NCR is the formal legally binding document that states we found a physical defect. Here is the scope of the problem and here is the approved protocol we are going to follow to fix it.
Emma: It is the absolute backbone of corrective action.
Ryan: But historically, generating an NCR is a tedious multi step administrative chore that everyone just hates doing.
Emma: Here’s where it gets really interesting. Consider the psychological reality of a manufacturing floor. You have an inspector Operating under intense time pressure, the assembly line is moving, raw product is piling up behind them. If the formal process to file an NCR requires them to, you know, back of their current screen, open three different software modules, manually retype long serial numbers,
Ryan: and search a directory for the correct standardized template.
Emma: Yes, what actually happens in reality, if
Ryan: we connect this to the bigger picture? Reducing friction isn’t just about saving clicks. Under that kind of intense pressure, the worker simply skips it.
Emma: They toss the defective part into a scrap bin, yell over to a mechanic to tweak the machine, and the formal documentation never takes place.
Ryan: Cumbersome bureaucratic processes actively breed non compliance.
Emma: So to dismantle those hurdles, version 7.26 introduces the frictionless one click NCR right
Ryan: from the screen of the failed test. An inspector can instantly generate a fully populated non conformance report with a single button press.
Emma: By stripping all the friction out of the reporting process, the software ensures that compliance actually happens.
Ryan: And they have deployed a bunch of new setup and compliance tools to support this. They built a dedicated NCR setup wizard complete with expanded sample data.
Emma: So new organizations can immediately see how a compliant workflow should function.
Ryan: Exactly. They also engineered an automatic numbering system designed explicitly to support strict reporting formats like Fair Form 3 or AS 9102.
Emma: And for anyone outside the defense sector, AS 9102 is a set of incredibly strict aerospace reporting standards. They dictate exactly how a report must be formatted and numbered for government auditors.
Ryan: And the auto numbering ensures that all that complex formatting is handled flawlessly by the back end. It doesn’t force a stressed inspector to remember a convoluted naming convention.
Emma: The release also attacks the severe bottleneck of getting approvals. They introduced native digital signatures directly onto
Ryan: the test records because previously if a process required a multi level approval chain, say the floor inspector signs off, then the QA manager, then the VP of Operations teams were forced to revert to paper dark ages. Truly, they would print a PDF of the failure, physically walk it around the facility to gather wet signatures, and then scan it back into the system.
Emma: Just describing that print, walk, sign and scan process feels exhausting.
Ryan: It is a massive waste of time. Now all of those stakeholders can sign off digitally right inside the database. It supports those complex approval chains without a single piece of paper changing hands.
Emma: And crucially, when the system finally generates the completed NCR document, it can now automatically embed original test PDF attachments directly into the generated NCR PDF.
Ryan: Which is brilliant.
Emma: So if our hypothetical worker Lisa in the paint department took a photograph of a deeply scratched metal surface. During the initial inspection, that exact photo
Ryan: is automatically stitched into the final PDF of the non conformance report. Nobody has to go hunting through shared network folders to compile a report for an auditor.
Emma: The document simply builds itself.
Ryan: It guarantees a watertight audit trail with zero extra administrative effort. And there are other great v7.26 features too, like an improved gate picker interface for faster data entry and a default
Emma: word report layout that natively supports batch reporting, meaning multiple tests in one report,
Ryan: which is another massive time saver. Imagine an auditor asks for the records of every test performed on a specific machine yesterday.
Emma: Without batch reporting, you’d be downloading 50 separate PDF files.
Ryan: Exactly. Now you can consolidate an entire day’s worth of tests into a single organized document with one click.
Emma: Okay, let’s synthesize the journey we’ve taken today. We started by looking at the incredibly fragile reality of physical clipboards and undocumented tribal knowledge.
Ryan: The overflowing filing cabinets.
Emma: Right. And we explored how Quality Inspector, which by the way is available globally through Insight Works network of over 750 partners, utilizes advanced Azure document intelligence to build a
Ryan: digital bridge capturing everything from a simple binary bicycle check to a highly complex pathogen test.
Emma: We saw how the engine automatically grades products and alerts management through teams. And then we analyzed how the v7.26 update takes the crucial step of turning a failed test directly into trackable pattern
Ryan: based action by implementing failure correlation analysis, enforcing tool calibration, and ruthlessly eliminating the friction of creating ncrs.
Emma: So what does this all mean for you? Whether you are manufacturing aerospace components, verifying food safety, or just trying to manage a team’s workflow, the ultimate lesson here is about human behavior.
Ryan: Absolutely.
Emma: The secret to continuous improvement is reducing the friction between finding a problem and reporting a problem. When a system makes it easier for a worker to do the right thing than to cut a corner, compliance becomes automatic.
Ryan: You stop fighting your own process and start actually improving the quality of your product.
Emma: But examining this evolution leaves us with a really fascinating final thought to mull over.
Ryan: Yes, v7.26 is all about finding correlations to spot patterns after the failures have already started happening. What does the future hold?
Emma: What’s the next frontier?
Ryan: Will the next iteration of quality control integrate AI to predict a batch failure before the raw materials even hit the shop floor?
Emma: Oh wow. Like an engine that analyzes ambient warehouse humidity, vendor history, and microscopic tool calibration drift all at once and instructs you
Ryan: to halt production before a single defective part is ever manufactured.
Emma: That is a profound thought to leave off on the definitive end of the reactionary era and the beginning of true prediction. We’ll leave you to explore that on your own. Thanks for joining us for this deep dive.
