Shop Floor Data Collection
You Can Trust.
Counts on a clipboard. Scrap in a spreadsheet. Labor when someone remembers.
We help manufacturers collect production data where it happens, connect it to the ERP, and turn it into analytics and forecasts your team will use.
Clean collection
beats clever analytics.
RADIN helps mid-market manufacturers build and protect their investment in AI with an independent AI readiness assessment before any deal is signed, an optional fractional Chief AI Officer, and the ERP and CRM systems to match, with no software quota behind the advice.
• HEAR IT FROM A CLIENT
See what our clients say
“RADIN stepped in and salvaged a very significant go-live date in two or three months, when we were at a failure point. That was massive for us as an organization. You said, listen, we can do this, and you’ve been able to do that.”
Rick Pace
HME. Three years working with RADIN.
What bad floor data costs you
Six problems that start with how the numbers get written down.
Counts get typed in at shift end. Good parts and scrap recorded at the machine.
Labor is a guess. Clock-ins against the job, so cost is real.
Two shifts, two definitions. One agreed meaning for a stop, a scrap part and a good part.
The report describes last month. Numbers reach the ERP the same day.
Purchasing buys just in case. Forecasts built on real order history.
Nobody trusts the dashboard. Analytics built on data that holds up.
Don't buy a dashboard until two shifts would record the same stop the same way.
Fix the definitions first. It costs almost nothing.
• THE GUIDE, BY PARKER HENDRICKS
Our view: don’t buy a dashboard until two shifts would record the same stop the same way. Clean collection beats clever analytics every time.
Where shop floor data comes from today
In most shops, production data is written down after the fact. Counts get typed in at the end of the shift, scrap goes on a clipboard, and labor gets estimated when someone remembers. IoT Analytics estimated in its 2025 MES market report that 54% of small and mid-size plants still run production on some mix of pen and paper or spreadsheets (IoT Analytics, December 2025).
Paper, spreadsheets, terminals and sensors
| Method | Up-front cost | Hidden cost | Good for |
|---|---|---|---|
| Paper travelers and clipboards | Very low | Retyping, missing entries, numbers that arrive days late | Very small shops and a starting point |
| Spreadsheets | Low | One person’s knowledge, version confusion, no link to the ERP | Short-term tracking on a single line |
| Shop floor terminals or tablets | Moderate | Training and keeping the screens simple | Labor by job, scrap, quality checks |
| Machine sensors and controller data | Higher | Setup and deciding what to keep | Counts, cycle time and downtime captured on their own |
The one thing to fix before buying anything
Definitions. Decide what counts as a good part, a scrap part and a stop, and make sure every shift records them the same way. Give parts, jobs and machines one naming scheme across the floor and the ERP. It costs almost nothing, and it decides whether every tool you buy afterward tells the truth.
What collected data is good for
Once the numbers hold, manufacturing data analytics answers questions you’ve been guessing at:
- What did this job really cost, and did we make money on it?
- Which machine or step is limiting output this month?
- Where is the scrap coming from, by part, machine and shift?
- What will demand look like next quarter, so purchasing can plan ahead?
That last one is where machine learning comes in. A forecasting model learns from your real order history. A failure model learns from machine signals and service records. Neither can learn from numbers that were guessed at the end of a shift.
Before you buy an analytics product
- Can two shifts record the same event the same way?
- Do floor data and ERP data use the same part and job numbers?
- Does the data reach the ERP the same day?
- Is there one question the tool must answer in its first 90 days?
If any answer is no, fix that first. We’ll help either way.
Where to start. Book a 30 minute call. Bring the workflow you want to fix, or the proposal someone just sent you. Your AI readiness assessment is included, and if your shop isn’t ready for AI yet, you’ll hear it from us first. Book a 30 minute call.
Collect it once.
Use it everywhere.
Once the numbers hold, the same data answers what a job really cost, where scrap comes from and what demand looks like next quarter. That’s where machine learning starts to pay.
Start where your shop is
A broken system needs the foundation first. Decent systems with no AI plan need the judgment. Most shops need both, in that order.
-
AI readiness assessment
Included when you work with us. Where AI fits on your floor and where it doesn't, before a dollar is spent. -
Fractional Chief AI Officer
$5,000 a month, 90 days then month to month. 16 hours, tracked. Any vendor quote answered inside 48 hours. -
CRM Health Check
Free. A current-state map, a 90-day plan and a one-pager for the board. Shops from $10M and 20 people. -
ERP for manufacturers
From $10M to $300M, with our own tool for the move off QuickBooks or a legacy system. -
CRM for manufacturers
Quoting and the customer file, connected to the floor and the books. Built only for manufacturers.
The numbers behind this
IoT Analytics estimates 54% of small and mid-size plants still run production on pen, paper or spreadsheets (2025). One in four US manufacturing workers is 55 or older (BLS, 2025).
Zero
software quota behind our advice
Start where your shop is
A broken system needs the foundation first. Decent systems with no AI plan need the judgment. Most shops need both, in that order.
-
AI readiness assessment
Included when you work with us. Where AI fits on your floor and where it doesn't, before a dollar is spent. -
Fractional Chief AI Officer
$5,000 a month, 90 days then month to month. 16 hours, tracked. Any vendor quote answered inside 48 hours. -
CRM Health Check
Free. A current-state map, a 90-day plan and a one-pager for the board. Shops from $10M and 20 people. -
ERP for manufacturers
From $10M to $300M, with our own tool for the move off QuickBooks or a legacy system. -
CRM for manufacturers
Quoting and the customer file, connected to the floor and the books. Built only for manufacturers.
The numbers behind this
IoT Analytics estimates 54% of small and mid-size plants still run production on pen, paper or spreadsheets (2025). One in four US manufacturing workers is 55 or older (BLS, 2025).
Zero
software quota behind our advice
Questions
Shop Floor Data: Common Questions
What is shop floor data collection?
It’s capturing what happens in production, like counts, scrap, labor, downtime and quality checks, where it happens, so the numbers reach your ERP without being retyped.
Paper, spreadsheets or sensors: which should we use?
Paper costs nothing up front and a lot later in errors. Spreadsheets help one person and break when they leave. Terminals and sensors cost more and pay back when the data feeds job costing and scheduling. Most shops end up with a mix.
What is manufacturing data analytics good for?
Knowing what a job really cost, where scrap comes from, which machines limit output, and what demand looks like next quarter. It only works if the data underneath it is collected the same way every shift.
What should we fix before buying an analytics product?
Make sure the same event gets recorded the same way on every shift, and that parts, jobs and machines follow one naming scheme. Analytics on inconsistent data gives confident wrong answers.
Do we have enough data for machine learning?
Often more than you think, but it’s rarely clean. We check what you have in the assessment and tell you whether a model would be reliable before anyone builds one.
• Built only for manufacturers
We know which floors
we belong on
Manufacturing of every kind, from job shops and fabricators to machinery builders and the suppliers feeding aerospace, defense, medical and electronics.
Who we work with
Manufacturers with 20 to 1,000 people and $10M+ in revenue, from founder-led job shops to larger manufacturing organizations. If aging or disconnected systems are holding you back, we help you plan and deliver the right AI, ERP, and CRM improvements.
Manufacturing
Know what a job really cost. See the floor without three phone calls.
• Built only for manufacturers
We know which floors
we belong on
Manufacturing of every kind, from job shops and fabricators to machinery builders and the suppliers feeding aerospace, defense, medical and electronics.
Who we work with
Manufacturers with 20 to 1,000 people and $10M+ in revenue, from founder-led job shops to larger manufacturing organizations. If aging or disconnected systems are holding you back, we help you plan and deliver the right AI, ERP, and CRM improvements.
Manufacturing
Know what a job really cost. See the floor without three phone calls.
People who have stood
where you stand
Parker Hendricks grew up on the floor of his father’s metal fabrication shop in Monroe, Georgia. He holds the AI seat at the Georgia Manufacturing Alliance and hosts The unManufactured Show.
Sixteen years of Dynamics delivery behind him, a hybrid US and offshore team, and one rule: we show up, we do what we say, and we don’t disappear after the handshake.
“We build on the Microsoft stack, and where it fits your shop we will scope it. So here is our rule: we disclose what we earn on anything we recommend.”
Parker Hendricks, VP of Sales
People who have stood
where you stand
Parker Hendricks grew up on the floor of his father’s metal fabrication shop in Monroe, Georgia. He holds the AI seat at the Georgia Manufacturing Alliance and hosts The unManufactured Show.
Sixteen years of Dynamics delivery behind him, a hybrid US and offshore team, and one rule: we show up, we do what we say, and we don’t disappear after the handshake.
“We build on the Microsoft stack, and where it fits your shop we will scope it. So here is our rule: we disclose what we earn on anything we recommend.”
Parker Hendricks, VP of Sales
Book a 30 minute call
Bring the clipboard, the spreadsheet or the report nobody trusts. We’ll show you what collecting it properly would take.
