Key takeaways
The night shift leaves a total on the whiteboard and a note: two stops, both mechanical. By the morning meeting nobody knows when they happened, how long they lasted or why the total fell short.
Maintenance remembers one stop; production remembers three.
Software that tracks production replaces the whiteboard with a record the line writes itself. This guide is for the plant manager, operations manager or CI lead who has to choose that system and get the floor to use it.
Production tracking software, also sold as shop floor tracking software or production reporting software, records what each machine and line does during the shift:
A stop on a bottling line, as it should be stored:
| Field | Example | Where it comes from |
|---|---|---|
| Machine and line | Filler, Line 2 | Set up once |
| Start and end | 10:42:15 to 10:49:40 | Machine signal |
| Duration | 7 min 25 s | Calculated |
| Reason | Material shortage | Operator, on a tablet |
| Product | 0.5 l bottle | Order or operator selection |
| Shift | B | Shift calendar |
Every report in this guide adds up rows like this one. If the rows are wrong or missing, no dashboard can repair them.
Manual tracking is a sensible start and costs almost nothing. It fails in three ways.
A whiteboard total tells you what happened once nothing can be changed, and a spreadsheet keyed in the next morning is older still.
That is the case for real-time production tracking software: a supervisor needs to see a stopped machine during the stop, while a response can still change the shift's output.
Hardly anyone writes down a 90-second jam. Yet ten of them make 15 minutes, and on a machine that makes one unit every six seconds, that is 150 units nobody can explain.
Never logged, those stops show up as slow running in the OEE, and the team hunts for a speed problem instead of a jam.
When production, maintenance and quality each keep a sheet, the meeting is spent deciding whose figure is right, and reasons filled in at shift end drift toward "other". Machine timestamps settle how long and how often, which moves the meeting on to why.
PLC and machine signals: Automated machines already know when they run, stop and finish a cycle. Tracking software reads those signals from the controller.
Add-on IoT sensors: Older machines with no usable signal get a small external sensor that detects each cycle, or whether the motor is running.
Cameras: Manual stations such as hand assembly or packing may give no signal at all. A camera running computer vision can tell when the line is standing still.
Agree camera use with staff representatives first.
The operator: No signal knows why a line stopped. A good system fills in the stop's start and length and asks the operator only for the reason, in a couple of taps on a tablet.
| Method | Pros | Cons |
|---|---|---|
| PLC or machine signal | Exact timestamps and counts, no extra hardware | Needs controller access and someone who knows the program |
| Add-on IoT sensor | Quick to fit on older machines, program untouched | Sees run, stop and cycles, not alarm codes |
| Camera | Works on manual lines with no signal, video of each stop | Needs a clear view and steady lighting |
| Operator input | The cause of each stop, plus rejects no sensor counts | Lapses if the reason list is long or the data goes unused |
Fabrico's machine monitoring software, for example, logs stops from all three machine sources.
Good production reporting software builds four views from the same tracked data, with nothing retyped.
Which machines are running, which are stopped and for how long, and output against target this hour, on a screen at the line and on the supervisor's phone.
Produced when the shift ends, not typed up afterwards: output, good and rejected units, OEE, the longest stops and any stop still missing a reason. The next shift reads it at handover.
Stop time ranked by reason over a week or a month, by minutes and by number of stops, because a hundred short jams and one long breakdown need different fixes.
Availability, performance and quality, comparable across lines, machines and shifts. Good availability with poor performance points to speed or short stops, not breakdowns.
Not sure you need OEE yet? Read production monitoring vs OEE.
These seven checks show how a system will behave on your floor.
Ask for a dated plan from signature to live counts on one line: hardware, network, PLC work and IT approval. If the first count is months away, the floor loses interest.
Ask to see your oldest machine and a manual station connected, not the newest one. If operators would type the numbers in, you are buying a digital whiteboard.
Hand the tablet to an operator during the demo. Tagging a stop should take seconds, with start time and duration filled in and reasons grouped in the operator's language.
Ask the vendor to walk through one shift's OEE: planned production time, planned stops, changeovers, the ideal cycle time per product and the short stop threshold. If two systems disagree on the same shift, one applies a rule you cannot see.
Our guide on how to calculate OEE lists the rules to agree first.
Orders, product data and standard rates usually live in the ERP. Check which way data flows, whether there is an API, and who maintains the connection when either side changes.
A breakdown stop is the start of a repair. Check whether a technician can open a work order from the stop itself, with machine and time attached, or must retype it into a separate system.
Ask where data is hosted, which certifications the vendor holds (for example ISO/IEC 27001 for information security), who sees what, and how you get your data back if you leave.
Need production scheduling too? Ask whether it ships today or sits on a roadmap.
For a pilot scorecard, see our comparison of real-time production monitoring software; for the wider field, our shop floor management software guide.
Choose the line that limits plant output, or the one with the loudest arguments about its numbers. Write down what success means first, such as counts that match a manual tally and a reason on every long stop.
Run the system next to the old sheet for a few shifts and compare totals. A count nobody trusts gets blamed for every bad number later.
Start short, split into planned and unplanned, in the words used on the floor. Add a reason only when "other" keeps growing for the same cause.
Take the whiteboard down and open the shift report and loss Pareto in the morning meeting. If the meeting still uses the old sheet, the pilot has not started.
Fix the top item on the Pareto and show the line its own before and after. Operators keep tagging reasons when they see the data used.
Reuse the same reasons, definitions and report layout on every new line, so OEE means the same thing everywhere.
Fabrico is our product, so weigh this section accordingly. It is a cloud-based platform that combines MES, live OEE and a built-in CMMS for maintenance.
A pilot site is operational in days; for OEE with hardware, connecting the machines takes about a month and the full rollout 3 to 4 months. Most customers see ROI in 3 to 6 months.
Fabrico is built in the EU, hosted on AWS in an EU region and certified to ISO/IEC 27001, ISO/IEC 20000-1 and ISO 9001. More on our OEE software and downtime tracking software pages.
Customer results:
What it does not do today: production planning and scheduling is on the roadmap, not in the product. Machines without a usable signal need a sensor or camera fitted first, so plan that into the pilot.
Want to see one of your own lines tracked live? Request a demo.
A system that logs each line's output, rejects and stops as they happen, with the start, duration and reason of every stop. That record feeds live screens, shift reports and OEE.
Tracking captures events: counts, stops and rejects with timestamps. Reporting turns them into shift reports, loss charts and OEE trends.
Use one system for both, so the reports match what happened on the floor.
Mostly. Some vendors mean following jobs and work in progress through each operation; production tracking usually means output and downtime by line and machine.
Ask which one a vendor means.
Fit an add-on IoT sensor that detects cycles or whether the machine is running. On manual lines with no machine signal, a camera can see when the line has stopped.
Operators add the reasons on a tablet.
Yes, but far less. Stop times come from the machine, so operators pick the reason and log any rejects no sensor counts.
Keep the reason list short and use the data daily, or the tagging fades.
With a usable PLC signal or a sensor, live counts can appear within days; trusting them takes a few weeks of checks. With Fabrico, a pilot site runs within days, and a full OEE rollout with hardware takes 3 to 4 months.
Login
Blog
