Objects

Two sensors, one puff, and a controller that retunes when nothing arrives

A new puff-sensing patent lands in a heated-tobacco business where control, patent ownership, and combustion are still contested.

Clay-diorama editorial illustration related to Aerosol-Generating Device With Improved User’S Puff Detection And Method For Improving A User’S Puff Detection.

What changes in this design

A reader might assume a device makes its puff decision from one reading and keeps its detection settings fixed.

Here, two components detect different puff-related parameters through different sensing methods. If a detected parameter indicates no puff, the controller adjusts sensitivity, meaning how readily detection responds, and/or resolution, meaning how finely it distinguishes readings.

It is worth knowing because a no-puff indication can cause the detection settings to change. That tells the reader the device reacts to that result rather than simply leaving the settings as they are.

The puffs that change the numbers

At 25 watts, with a 1.2-ohm coil, more than 90 percent of test puffs were dry puffs across every flavor tested. In laboratory work on e-cigarette emissions, readings for formaldehyde, acetaldehyde, and acrolein ran 30 to 250 times higher than under normal puff conditions. The number that stopped me was not the wattage. It was how quickly an ordinary-looking use could turn a measurement problem into a very different exposure problem.

That study does not test the product in this filing, and it does not prove what any heated-tobacco device emits. It does establish the stakes of recognizing a puff properly. A device that misses a draw, or mistakes one for another, is not merely losing a convenience feature. It may make its next heating decision on the wrong version of what the user did.

Two signals, one decision

US20260256210A1 arrives precisely in that narrow, consequential gap. Its device uses a first and a second component to detect different puff-indicative parameters through different modalities. When a detected parameter indicates no puff, a controller adjusts sensitivity and/or resolution. It creates a feedback loop after a no-puff indication.

That design problem has a physical history. A 2019-priority patent family assigned to Philip Morris Products S.A. explains that sensors placed inside the article cavity can restrict airflow and can be damaged during cleaning or while an article is inserted or removed. Another earlier filing used a thermopile mass-airflow sensor rather than a pressure sensor and contemplated heating in proportion to draw strength. The sensing method itself has become territory to own.

Nine countries and a sensor

That ownership question has already carried remarkable legal weight. Philip Morris began an action against BAT in Japan in 2018; BAT responded in 2020 with cases at a U.S. trade body and across Europe. The dispute eventually spanned nine countries. A global settlement announced on February 2, 2024, provided for an eight-year term and resolved the outstanding injunctions, including a previous U.S. import ban on IQOS.

This dispute is not confined to old court files. A related Philip Morris application, published in 2025, selects an operating profile from puff behavior. In February 2026, the company's UK and Ireland announcement for IQOS ILUMA i ONE described a Flex Puff feature that adapts device behavior to the user's puffing rhythm. The announcement does not disclose its sensing hardware, so it cannot show whether this filing supplies that feature. It does show that puff adaptation is now a selling point.

Heat inside the stick

The commercial and regulatory context makes the precision feel less incidental. By the end of 2023, smoke-free products accounted for 40 percent of Philip Morris revenue, according to a public IQOS summary. Yet a study using low-cost thermometers, carbon monoxide and particulate sensors on an IQOS ILUMA recorded 284 degrees C inside the stick, 300 degrees C at its heating blade, carbon monoxide rising to 43 ppm, and PM2.5 up to 887 micrograms per cubic metre in a use. Its authors described the remaining tobacco as charred and concluded that calling the process non-burning was a fallacy.

Those results are one study, not a final verdict on the whole category. A cleaner control loop does not resolve them. The FDA's 2019 U.S. authorization of IQOS through the PMTA pathway cited substantially lower acrolein and formaldehyde levels than combustible cigarette smoke, while withholding reduced-exposure marketing claims pending separate review. The decision was conditional, not a categorical safety finding.

The open question

Multi-sensor measurement looks sensible until a person's draw falls outside the conditions anticipated by the calibration. An electronic-cigarette study pairing an optical sensor with a pressure sensor reported about 5.6 percent overall relative error, but error rose to 11 to 17 percent when the optical signal exceeded its calibrated range and 9 to 13 percent at inhalation pressures of 300 to 600 Pa. I cannot tell from US20260256210A1 whether its two modalities avoid those edge failures. The publication's stated proposition is adjustment, rather than a reported error rate.

That is why I read this filing as evidence of an industry still trying to resolve a narrow but consequential sequence: a breath, a heater, and a decision between them. Better detection could reduce bad control decisions and may improve the experience the device is designed to deliver. My concern is the larger claim that such precision settles what the device is doing to tobacco or to people. It does not. The most important question is not who can sense a puff most elegantly, but whether the system being optimized has earned the confidence placed in it.

Sources

Reporting for this piece drew on the published filing and the references below.

The record, briefly. US20260256210A1, “Aerosol-Generating Device With Improved User’S Puff Detection And Method For Improving A User’S Puff Detection” An aerosol-generating device includes first and second components that detect different parameters indicative of a user’s puff using different modalities, and a controller adjusts sensitivity and/or resolution when a detected parameter indicates no puff.

More angles on this

CPC A24F: Smokers' requisites; match boxes; simulated smoking devicesThe official classification family this record was filed into, published in the USPTO Cooperative Patent Classification scheme.
A Baseplate's Signal: From Guess to Defined ResponseAnother reading from the neighbouring A24B classification family, manufacture or preparation of tobacco for smoking or chewing; tobacco; snuff.
Degassing a bottle: a chamber, a pump, and the judgment left to whoever holds itAnother reading from the wider A23L classification family, foods, foodstuffs or non-alcoholic beverages, not otherwise provided for; preparation or treatment thereof.
Small Game Cleaning Device Makes the Next Move Easier to SeeAnother reading in the same Objects category on this site.

Read around this one:Public patent record USPTO published figure