Most weeks, I get a similar call: someone says they need “a gas sensor”, sometimes even naming a model they found online, and asks for a quote. That sounds straightforward enough, but nine times out of ten, my first question is not about the sensor. It is about what they actually need it to do.

Same sensor, different problem. Two customers can ask for the same sensor but need very different solutions. One may be monitoring a confined space in a wastewater treatment plant; another may be working in a pharmaceutical cleanroom. The gas, and even the concentration range, might be the same, but the environment, level of risk, regulatory pressure, and consequences of a wrong reading can be completely different.

Why that matters. If I took those requests at face value, I would be doing the customer a disservice. This article explains why application context matters so much, what years of conversations across different industries have taught me, and how to approach sensor selection so you do not end up with the wrong tool for the job.

Why the Application Comes Before the Technology

Start with the application, not the sensor. It is tempting to begin with the sensor itself: the technology, the specification sheet, and the price. I understand why. However, the most useful sensor-selection conversations I have had always start somewhere else: with the application.

Key questions to answer first. Before discussing technology, it is important to understand:

• What is actually being monitored? Is it a specific gas, general air quality, particulates, or oxygen levels? The answer shapes everything that follows.
• Why is it being monitored? Personal safety, process control, environmental compliance, and product quality all place different demands on a sensor, even when the target substance is the same.
• Where will it be used? Indoors or outdoors, fixed or portable, confined space or open plant — location can significantly affect housing, power, and communication requirements.
• What will the environment do to it? Consider temperature swings, humidity, dust, vibration, corrosive atmospheres, and other gases that could interfere with the reading.
• How accurate does it need to be? More precision usually increases cost and maintenance. It is worth asking whether that extra precision would change any decision you would actually make.
• What operating life is expected? A sensor intended for a five-year unmanned installation needs a different design approach from one recalibrated weekly by an on-site technician.
• What maintenance can realistically be supported? Even the best sensor is only as reliable as the calibration and servicing regime behind it.
• Are there compliance expectations? I would always point customers towards the relevant guidance for their sector, but even knowing that a compliance framework exists changes how a sensor should be specified and documented.

Work through the context first. Once these questions are answered, the right sensor often becomes much easier to identify. Skip them, and you are effectively specifying blind, hoping the datasheet covers a situation no one has properly described to the supplier.

What This Looks Like Across Different Industries

The differences become clearest when you look at how the same underlying question, “how do I detect this gas reliably?”, plays out across sectors.

• Manufacturing. The challenge is often less about an unusual hazard and more about a demanding operating environment. Vibration, temperature changes near machinery, and dust build-up can all affect performance over time. It is also worth being specific: “industrial rated” is not a single, universal category. A metal-welding facility and a chemical process line may both sit under the manufacturing label, but they expose sensors to very different conditions. The sensor should be chosen for the actual floor environment, not just the industry name on the door.
  
• Oil & Gas. Some of the highest-consequence monitoring anywhere, flammable and toxic gases, explosive atmospheres, remote or offshore sites where a technician can't just pop out to recalibrate something. Reliability and hazardous-area certification tend to dominate here, often more than raw sensitivity. My advice: think hard about access. A technically excellent sensor that's awkward to service remotely causes more problems than it solves.
• Pharmaceutical. Cleanrooms bring strict hygiene requirements, low-particulate materials, and processes sensitive to any contamination the monitoring equipment might introduce. Accuracy and traceability matter enormously here, since the data often needs to stand up to audit. People sometimes assume a clean environment is an easy one to work, in practice the material and documentation requirements around a cleanroom sensor can be tougher than almost anywhere else.
• Food & Beverage. Similar hygiene pressures to pharma, but different gases, carbon dioxide in brewing and packaging, ammonia in refrigeration, oxygen displacement in confined storage. Wash-down cycles are a real practical headache: sensors need to survive routine cleaning without losing accuracy.
• Water & Wastewater. Confined space entry is a constant theme, chambers and below-ground access points where multiple gases (hydrogen sulphide, methane, oxygen depletion) can appear together. Humidity is the quiet enemy here; a sensor that performs beautifully in a dry lab test can struggle in a genuinely wet chamber. Buyers often underestimate how much real-world environmental testing matters versus lab-condition specs.
• Construction. Monitoring here is frequently temporary and mobile, moving site to site, exposed to weather and rough handling a fixed installation would never see. Robustness and ease of redeployment usually outweigh sophistication. A sensor perfectly specified for a permanent installation can be entirely wrong for a project that packs up every few weeks.
• Utilities. Long-term, often unattended installations across a wide geographic footprint. Battery life, remote communication, and low-maintenance design tend to matter as much as raw detection performance.
• Environmental Monitoring. Outdoor exposure, weather resilience, and often very low target concentrations that need distinguishing from background noise. Long-term data consistency usually matters more than instantaneous precision, you're building a trend, not taking a single reading.
• Laboratories. Controlled conditions, but very specific gases and tight concentration ranges tied to a particular process. Cross-sensitivity, a sensor reacting to a gas it wasn't designed for, is a common, under-appreciated risk, especially where multiple processes run side by side.
• Mining. Where relevant, this brings together some of the harshest conditions of any sector, dust, vibration, humidity, and genuinely life-critical detection, often underground with limited communication and access. A good example of where “good enough” specifications simply aren't.

What Years of These Conversations Have Taught Me

Three lessons keep coming up. Regardless of industry, the same themes appear again and again.

Similar problems can cross industry boundaries. A confined-space issue in wastewater may have more in common with a confined-space issue on a construction site than with a similar-looking gas detection task in a completely different environment. The label on the sector matters less than the conditions around the sensor.

Good questions lead to better outcomes. Asking the right questions usually matters more than recommending a product straight away. In my experience, the best results come when customers are willing to discuss the application properly before moving on to a specific model.

Start with the problem, not the catalogue. That is probably the most useful point I can pass on. I have lost count of the times someone has been convinced they need a particular sensor, only for the conversation to reveal that their actual monitoring objective calls for something quite different.

Practical Takeaways

If you're heading into a sensor selection process, here's what I'd actually suggest:

• Write down the application before you look at any spec sheet. What, why, where, and under what conditions.
• Be ready to answer questions a supplier should be asking you. If they're not asking about your environment, that's worth noticing.
• Don't assume “industrial grade” means suitable for your specific industrial environment. Ask what it was actually tested against.
• Think about maintenance access early, not after installation.
• Have an honest conversation about how much accuracy you actually need — and what decision that accuracy is meant to support.
• Avoid choosing a sensor based purely on a model number you've seen used elsewhere. What worked for someone else's application may be entirely wrong for yours.

It's easy to skip all this when you're under time pressure and just want a quote. In my experience, the ten minutes spent on these questions upfront save far more time, and money, than they cost.

A Final Thought

Sensor technology has advanced significantly, and capable hardware is rarely in short supply. What determines whether a project succeeds is not the spec sheet alone, but whether the application has been understood well enough to choose the right solution. The specification only matters once that groundwork has been done.

I would be interested to hear how this has played out for others. Have you seen a sensor selection go wrong because the application was not fully understood at the start, or found that the “obvious” choice turned out to be the wrong one? If so, I would welcome your thoughts in the comments.