5 data areas you should already be tracking on your production lines

Let’s start with energy. Not because it’s trendy, but because it hits your bottom line and your ESG requirements in one shot.

If you're still relying on your monthly utility bill and one big main meter to tell you how you're doing, you may not be missing the forest, but you’re surely missing every single tree.

“Energy is many things,” Stefan Jensen explains. “For most people, it’s kilowatt-hours or CO₂ emissions. But energy can also be temperature, humidity, pressure, flow. All can be data points that make sure your production is running as it should.”

The secret here is simpler than you might expect: setting up sub-meters. With sub-meters, you get visibility into where energy is actually going, whether it’s specific lines, zones, or even individual machines. And while it may seem too simple, sub-metering lets you isolate energy hogs, spot idle consumption, and in the end, help you set real KPIs.

Beyond just power, production environments also need to monitor related metrics like air flow, cooling systems, even compressed air usage. Flow sensors, thermal sensors, and pressure transmitters are your friends here. They’re cheap insurance against waste and a strong foundation for carbon reporting.

As Stefan Jensen adds, more customers will be asking for your energy footprint per part, and the hard truth is, that if you can’t give them real numbers, someone else will.

Every Head of Production we talk to says the same thing: “Our lines are running well.” But when we look at data like actual uptime, cycle times, target vs. real output, it’s usually a different story. And needless to say, you can't improve what you're not measuring (and no, “well” isn’t a metric).

“A good place to start is simply understanding how much of the time your machines are actually running,” Stefan Jensen says. “Then you can look deeper into why they stop, and whether you’re hitting your expected output.”

So where do you start?

The basics are simple: counter sensors like photoelectric sensors for part detection to determine machine cycle time, and IO-Link capable sensors to track states in real time. With those, you can monitor uptime, cycle speed, and transitions automatically.

Then go deeper: When you see downtime, why did it happen? Was it a jam? A material delay? Operator absence? You can track many of these manually at first, but eventually you’ll want sensors or software that track causes as you go. Integrate with your MES or use even a simple tablet interface for operators to log root causes in real time.

With a handful of simple sensors, you’ll get the data you need to begin securing a profitable production.

And once you’ve begun, next step is to measure environmental factors in your production where unexpected humidity or ambient temperature changes can subtly slow a machine down. This is where environmental sensors like temp/humidity monitors come into play – and not just for compliance but for performance as well.

Bottom line: efficient production is about visibility of your production metrics. Without it concrete datapoints, you’re planning in the blind.

This is a big one. Quality is the difference between a part that ships and a part that gets scrapped. But it’s also the difference between a smooth audit and a total compliance headache. And you’ve got to handle both.

“I like to split quality into two buckets,” Stefan Jensen says. “There’s product quality; does the part look and perform right? And then there’s compliance or process quality; are we sticking to the parameters we promised?”

On the product side, things have gotten both better and cheaper. Vision sensors and smart cameras can now detect surface defects, color mismatches, misaligned components, and missing parts. OMRON’s FH Series vision systems, for example, can check features at high speed inline, without slowing down your process.

Combine those with displacement sensors for precision measurement and color sensors to verify labeling or packaging, and suddenly you’ve got a reliable quality gate that doesn’t rely on a tired operator with a clipboard.

On the compliance and process quality side, you’re often dealing with standards, tolerances, and audits: ISO 9001, FSSC 22000, you name it. Whether it’s a required oven temperature, cleanroom humidity, or pressure differential, those numbers matter.

Again, automated process and sensors are the clever choice, Stefan Jensen explains.

“You do not need five people collecting data manually. You can collect it once, store it right, and only act if it’s out of tolerance.”

Environmental sensors, thermocouples, and barometric pressure sensors can all feed this kind of data straight into your dashboard, and give you those compliance reports you need, while allowing you to only act on anomalities, alerts, and out-of-the-ordinary events.

You’ve probably got a maintenance schedule on the wall. Machines get checked every X weeks, just like the OEM manual said. And most of the time, that works fine. Until it doesn’t.

“We see a shift toward using data for maintenance,” says Stefan Jensen. “Not just time-based, but based on actual running hours, vibration levels, power draw. All are signs that something might be going off.”

This is where vibration sensors, current monitors, and infrared temperature sensors become critical. Say a motor starts drawing 10% more current than usual. That’s your early warning. Or a gearbox starts vibrating just outside the norm. You’ve got time to act before it throws the belt and takes your whole line down, but only if you notice it.

In practice, you don’t have to ditch calendar-based maintenance but just layer in real-time machine condition data so you can prioritize. If a low-usage machine is still in perfect health, skip the full teardown. If a high-run robot starts showing signs of strain, bump it up the list.

The goal isn’t fewer maintenance tasks, but maintaining you equipment on a smarter schedule that’ll keep production running smoothly, all the time.

Here’s where it all ties together. If you’re not using production data to calculate real costs (per order, per line, per shift), you’re just crossing your fingers and hoping your gross margin doesn’t disappear.

“With data, you can now calculate the real cost per part or per order,” Stefan Jensen says. “How long did it actually take? How much energy did we use? How many people were involved? It all adds up.”

Start simple by tracking order start and stop times with basic input sensors or scan events. Then layer in energy use (from your sub-meters), labor time (from operator logins or presence sensors), and detract the scrap rates. and voila!, your costing is no more based on assumptions, but on actual, countable facts taken from you very own production line.

And here's the part no one talks about: once you have this data, you can decide what not to produce. Which low-margin orders are sucking up time? Which machines are costing you more per unit than planned? This is the kind of insight that separates good plants from great ones.

Don’t fret. You don’t need to become a data scientist overnight. But if you’re not tracking these five areas – energy, efficiency, quality, maintenance, and cost –you're playing defense. The systems are out there. The sensors are cheap. The integration is easier than ever.

As Stefan Jensen puts it: “The tools are there. The mindset is shifting. Now it’s just about acting.”

Pick the one area where you’re bleeding the most and start there. Build trust in the data and keep going. Because while your lines might still be running fine today, someone else is already using data to run faster, cheaper, and smarter, and will wait for you at the finish line, if you don’t keep up.