Not legally — but commercially it is very difficult to compete at the top of the market without it, because the sensing element sets the detection limit and the certification attaches to the plant that made it.
The sensing element is the product. A gas detector's usefulness is decided by whether its sensor responds to the target gas at the concentration that matters, how it behaves when exposed to interfering compounds, and how long it holds calibration. Those are properties of chemistry and materials, not of the enclosure or the electronics. A manufacturer that buys electrochemical cells or infrared optical benches inherits whatever drift, poisoning behaviour and replacement interval its supplier offers, and cannot improve them by redesigning the instrument around them. This is why Riken Keiki states it develops in-house every sensor at the heart of its detectors, which is what allows it to sell into semiconductor fabs that must detect hydride gases at part-per-billion concentrations, and why MSA Safety manufactures its own XCell sensing elements rather than buying them.
The Owned Hazard Sensor Model scores four weighted dimensions:
• Owned Production Scale (35%) — number and physical scale of manufacturing plants under direct control, the share of safety-critical components produced in-house, and automated production and calibration capacity.
• Safety Category Production Share (26%) — the proportion of total output and revenue devoted to safety detection instruments.
• Global Revenue and Service Network (21%) — audited group revenue, revenue attributable to safety detection, and the density of the owned service and calibration network.
• Certification and Manufacturing Credibility (18%) — hazardous-area approvals held by the manufacturing entity, standing among safety engineers, and the manufacturer's own recorded safety performance.
The 35% weighting on owned production reflects a regulatory reality, not a preference. A detector sold into an explosive atmosphere carries ATEX or IECEx approval, and that approval is granted to a manufacturing process at a named facility operating under a controlled quality system. Contract manufacturing changes the process, which changes what was approved. A brand owner that outsources production can still sell instruments, but it cannot hold the approval in its own name in the way a manufacturer does.
Vertical integration is not automatically better engineering. Several excellent detectors are assembled from bought-in sensors, and owning a sensor plant does not guarantee a superior instrument. What ownership provides is the ability to move the specification — to chase a lower detection limit, a longer calibration interval or a resistance to a newly identified poison — without waiting for a supplier to decide the market justifies the investment. At the top end of the market, where the customer is a semiconductor fab or a nuclear operator, that ability is usually the difference between winning and losing the specification.
Disclaimer: This ranking is compiled from third-party authoritative sources including audited financial statements, stock exchange disclosures and regulatory registries. VerityRank is independent and receives no compensation from any company for inclusion, exclusion or position.
Because ionising radiation cannot be detected chemically. It has to be converted into a measurable signal by interacting with matter, and the material that does the converting — a crystal, a gas or a semiconductor — determines what the instrument can see.
The detection problem is fundamentally different from gas detection. A gas sensor measures a chemical concentration by allowing molecules to react at an electrode or absorb infrared light. Radiation is not a chemical: it is energy travelling as particles or photons, and it must be stopped by something that converts that energy into electrical charge. The three practical approaches are gas-filled detectors such as Geiger-Müller tubes and ionisation chambers, semiconductor detectors, and scintillation detectors in which a crystal absorbs radiation and emits a flash of light that a photodetector converts into a pulse.
What the crystal decides is whether the instrument can identify a source, not merely detect one. A Geiger counter tells you that radiation is present and roughly how much. It cannot tell you whether you are standing next to medical isotope, industrial caesium-137 or naturally occurring potassium — and for a border monitor, a waste handler or an emergency responder, that distinction is the entire requirement. Distinguishing isotopes requires gamma spectroscopy, which needs a detector with enough energy resolution to separate gamma peaks that sit close together, and that resolution is set by the purity and crystal structure of the detector material. High-purity germanium offers the best resolution available and is exceptionally difficult to produce, which is why Mirion grows and processes its own germanium and scintillation crystals rather than buying finished detectors. Owning that process is what allows it to set its own energy-resolution specifications.
Different hazards need different materials, and no single detector covers all of them. Gamma and X-rays are detected with germanium, sodium iodide or caesium iodide; alpha and beta contamination requires thin-window detectors because those particles are stopped by air or a few micrometres of material; neutron radiation, which surrounds reactor cores and spent fuel, interacts strongly with materials such as helium-3 or boron and is effectively invisible to a gamma instrument. A nuclear site therefore operates a portfolio of detectors, each matched to a radiation type, which is why the instrument count on a reactor site runs into the thousands and why service and calibration infrastructure matters as much as the hardware.
Why the market structure differs from gas detection. Because radiation monitoring is driven by nuclear regulators rather than by plant operators, the specification is unusually rigid: an instrument either holds the required approval and traceability to a national standards laboratory or it cannot be installed. That rigidity favours a small number of material owners over many instrument assemblers — which is why the radiation segment of this ranking contains a focused specialist such as Mirion alongside a diversified scientific instrument group such as Thermo Fisher Scientific, and why no gas detection specialist competes there.
They answer to different regulators, sell into different buying cycles, and detect fundamentally different things — which is why a group that owns both treats them as separate businesses rather than as one product line.
Gas detection is occupational safety; fire detection is building safety. A gas detector protects the people working in a plant from an atmosphere that has become toxic or explosive, and it is specified by the plant's process safety engineers under industrial standards. A fire detector protects the occupants of a building from a fire that has already started, and it is specified by the building's designers under construction and fire codes. The buyer, the regulatory framework and the installation moment are different: gas detectors are procured during plant construction or upgrade, while fire detectors are procured with the building and then replaced on a replacement cycle driven by the building's compliance regime.
The technologies share almost nothing. Gas detection relies on electrochemical, catalytic, infrared or semiconductor sensors that respond to a chemical concentration. Fire detection relies on optical scattering chambers that see smoke particles, heat sensors, flame detectors that look for the ultraviolet or infrared signature of combustion, and increasingly multi-criteria detectors that combine several signals to suppress false alarms. A smoke detector has no sensor chemistry in common with a methane detector; the manufacturing processes, calibration equipment and certifications are entirely separate.
Halma's structure shows why the distinction matters commercially. The group owns both Crowcon Detection Instruments, a gas detection manufacturer at Abingdon, and Apollo Fire Detectors, a fire detection manufacturer at Havant — but they sit in different Halma sectors, with Crowcon in Environmental & Analysis rather than in Safety. That separation is not administrative tidiness: fire detection demand follows construction activity and building replacement cycles, while gas detection demand follows industrial output and process safety regulation, and the two do not rise and fall together. Owning both hedges the group's exposure; running them as one business would blur two different sales channels and two different certification regimes.
For buyers the practical consequence is that an integrated safety specification rarely comes from one supplier. A plant typically buys gas detection from a specialist such as MSA Safety, Dräger or Honeywell, and fire detection from a building systems provider — and increasingly, from a manufacturer that has bought its way into both, as MSA did by acquiring Autronica Fire and Security for approximately USD 555 million in July 2026 to add maritime and critical-infrastructure fire and gas systems. The convergence is happening through acquisition rather than through product development, because the two businesses have little engineering in common.
By submitting the plant itself to audit. In hazardous-area and nuclear certification, a third party inspects the factory, the process and the records — so the manufacturer's quality system is what is actually being approved.
Product testing is only half of it. A detector has to pass performance tests: response time, accuracy, temperature behaviour, resistance to poisoning, ingress protection. But ATEX, IECEx and equivalent approvals also require quality-system audit of the manufacturing site, because a certificate is a claim about every unit produced, not about the sample submitted. That means the manufacturer must be able to demonstrate traceability of components, calibrated test equipment, controlled procedures and records that survive inspection years later. The practical consequence is that changing a plant, a subcontractor or a critical component can invalidate the approval — which is precisely why contract manufacturing is structurally difficult in this market.
The calibration chain is the second proof. A detector reads a concentration, and that number is only meaningful if the instrument has been checked against a known reference gas of certified composition, traceable to a national standards laboratory. Manufacturers that produce their own calibration gases, reference standards and test equipment control that chain internally; companies that buy it depend on a supplier's certificate. Dräger is the clearest example — its Dräger-Tubes are both a product and a reference method, and the company manufactures the colourimetric reagent chemistry and supplies the hand pump that makes the measurement reproducible. MSA Safety similarly supplies calibration gases and docking stations that test its own instruments automatically.
Automated test infrastructure has become part of the manufacturing evidence. A plant that bump-tests instruments on an automated docking station produces a machine-generated record for every unit, which is far stronger evidence than a technician's signature. Industrial Scientific's DSX docking stations and iNet platform exist for that reason: the test result is captured at the point of production or use rather than transcribed later. The same logic applies in radiation monitoring, where instruments must be traceable to a national standards laboratory and are recalibrated on a defined schedule with regulator-auditable records.
What a manufacturer's own safety record adds. There is no certification requirement that a safety equipment maker avoid injuring its own workforce, but it is the most direct evidence available about whether the company's process control is real. MSA Safety completed 2025 with zero lost time incidents across more than ten million hours worked, a figure it reports in its own regulatory filings — a claim a company with weak manufacturing discipline could not credibly make.
Because it converts a manufacturing decision into a legal one. The certificate names the plant, the process and the component, so a manufacturer cannot change any of them without re-qualifying the product.
Approvals attach to physical arrangements, not to designs. When a detector is certified for use in an explosive atmosphere, the approval covers the specific production process at the named facility — the materials used, the assembly sequence, the test regime and the quality system around them. Move production to a lower-cost plant and the certificate does not travel with the design; it has to be obtained again. That is why Honeywell assembles BW series monitors and XNX transmitters at a facility in Dammam, Saudi Arabia under a local programme rather than simply shipping certified products in, and why MSA Safety, Dräger, Teledyne and Halma have each built or expanded production in China to serve Chinese demand: a plant inside the market shortens the path to local approval and to the customer's delivery requirement.
Certification also constrains component substitution. A gas sensor is a certified safety component in its own right in many jurisdictions, and swapping a supplier's cell for a cheaper equivalent is not a purchasing decision — it invalidates the type approval unless the new component is covered. This is the mechanism that makes vertical integration so valuable in this industry: Riken Keiki developing all its own sensors, MSA Safety building its own XCell elements and Hanwei packaging its own measurement dies all mean the critical component is inside the certificate holder's own quality system rather than outside it.
Nuclear monitoring adds a second layer. Radiation instruments must be traceable to a national standards laboratory and recalibrated on defined intervals with records a regulator can audit, which is why Mirion grows its own germanium and scintillation crystals: a spectrometer's energy resolution is set by material purity, and a change in crystal source changes what the instrument can legitimately claim to identify. Thermo Fisher operates in the same segment with a similar dependence on controlled materials and traceable calibration.
The commercial consequence is that certification is a moat that protects manufacturers and constrains buyers simultaneously. It keeps credible suppliers in the market by making entry expensive, and it prevents customers from switching to a cheaper detector whose certificate does not cover their installation. Manufacturers that hold approvals across multiple jurisdictions — Dräger with German engineering and global type approvals, MSA Safety with approximately 14 primary manufacturing locations across eight countries, Honeywell with more than fifteen detector plants — can serve a multinational customer from inside several regulatory regimes at once. That breadth, not unit cost, is what wins global safety specifications.
