A strap-on temperature probe, a current clamp, and a couple of air sensors can report most of what matters about an HVAC or refrigeration system's health, continuously, from outside the refrigerant circuit. US federal law is a big part of why that combination wins at scale.
EPA Section 608 requires certification for anyone who could be reasonably expected to violate the integrity of the refrigerant circuit. The list of covered activities includes the obvious ones: adding or removing refrigerant, replacing components, cutting lines. It also covers plain measurement. EPA's definition of a technician explicitly covers attaching and detaching hoses and gauges to and from an appliance to measure pressure.
That is reading a pressure, not changing anything. In the eyes of the EPA, putting gauges on a system is technician work, and the rule carries five-figure federal penalties per day, per violation.
For equipment monitoring at scale, that rule splits sensors into two categories. A sensor that lives inside the circuit, like a pressure transducer tapped into a service port, is legally a certified-technician install, per unit, per site. A sensor that stays outside it, a strap-on temperature probe or a current clamp, can be installed by facilities staff, an electrician, or the operator themselves, with no certification anywhere in the process. When you are looking at 40 rooftop units across a dozen sites, that difference decides what monitoring costs and how fast it deploys.
The definition is simple: the refrigerant-carrying parts of the system are never physically breached, so there is no risk of losing refrigerant.
No breach also means no new leak point.
Current signatures. A clamp around the supply conductor never touches the machine, and it sees more than power consumption. Research on reciprocating compressors has shown current analysis can detect and differentiate internal faults including valve leakage, a refrigerant-side problem read entirely from the electrical side. At field level, the useful signals are simpler: start current drifting up against the unit's own baseline, running current creeping as a condenser fouls, and the timing of trips.
Cycling patterns. Starts per hour, cycle length, duty cycle. These are simple counts. A healthy system cycles roughly 3 to 5 times per hour; short cycling compresses that to 10 to 20. Continuous monitoring adds the history: you see the count climbing week over week instead of discovering it after the compressor fails.
Surface temperatures. A single strap-on sensor on the discharge line can detect low-side faults, loss of charge, blower failure, a restricted metering device, because the discharge line concentrates compression heat. Purdue's virtual sensor research showed that the five key pressures of a vapor-compression cycle can be estimated from low-cost surface-mounted temperature sensors. Those are estimates from lab-validated models on fixed-speed equipment, not measurements.
Air-side deltas and box behavior. Supply and return air temperatures confirm whether the unit is actually delivering conditioned air, which is the outcome the building cares about. On refrigeration, box behavior adds more: recovery time after a door-open or a defrost, widening temperature swings, defrost cycles losing their regularity.
Direct pressure readings stay invasive. Textbook superheat and subcooling stay invasive too, because both require saturation temperature, which comes from pressure. And every repair is invasive by definition, and EPA-regulated. SmartHVAC's non-invasive monitoring guide walks through the full line between what external sensing infers and what stays gauge work.
That defines what non-invasive monitoring is for. It detects that something is wrong, which family of fault it likely is, and how fast it is moving, then flags it for a technician rather than diagnosing it.
Retrofit monitoring products are non-invasive by design.
The in-circuit readings live with OEM integration. Copeland's CoreSense uses the compressor as a sensor, with protection and diagnostics for low oil pressure, excessive discharge temperature, and high discharge or low suction pressure, delivered by an electronic module on Copeland compressors. They also ship with the compressor, so they cover the equipment they arrive on and nothing else.
The trade-off is coverage. Non-invasive monitoring covers every machine you already own. A typical operator's estate is mixed-brand by accident of history, rooftop units from different years, walk-ins from different vendors, and no single OEM's diagnostics reach across it. Watching all of it is a fleet management problem as much as a sensing one, and for a mixed fleet, covering everything matters more than depth on any single unit.
The certified technician still confirms and fixes every fault. What monitoring changes is the visit itself. Instead of rolling a truck to "unit not cooling" and starting diagnosis from zero, the tech arrives knowing which unit, which family of fault, and how long it has been developing, with days of trend data behind the dispatch. The regulated instrument stays in certified hands, where the law puts it, and the tech starts the visit already knowing where to look.
This is the approach SmartHVAC is built around: fleet-wide, brand-agnostic monitoring from outside the circuit, with retrofit sensors that clip onto the refrigerant lines and never open the system. It runs on Blynk's ready-to-use, scalable cloud infrastructure, the same platform HVAC manufacturers like Raypak and Windmill build their connected products on.
The platform layer is what turns sensor readings into a fleet view. Native iOS and Android apps put the trend data in the technician's hand, with white-label apps at Enterprise tier for OEMs shipping under their own brand. The retrofit hardware connects through Blynk's device SDKs and takes firmware updates over the air, while Data Converters bring in equipment that already reports data, no firmware changes needed.
.png)
