An Indonesian product recall has uncovered a problem in air monitors used to track smoke from peat fires: when humidity rose above 82%, the devices often reported far less fine-particle pollution than was in the air. Tests of 312 units found a median undercount of 39%, according to researchers who examined the devices after their manufacturer announced a recall on 14 October.
The monitors fed readings to the Kalimantan Air Watch public dashboard and to local warning teams in Central Kalimantan and South Sumatra. Researchers compared archived readings from 2024 and 2025 with reference instruments, finding 11 pollution episodes when those instruments measured more than 100 micrograms of fine particles per cubic metre for six consecutive hours, but the dashboard showed peaks below 55.
A flaw in the field network
The AwanTerra M2 was marketed as a portable monitor for community stations and local authorities. The recall covers 18,600 units sold since 2022. A notice from the fictional Directorate for Product Standards says a humidity-compensation fault may produce “materially low” readings in damp conditions; owners have been told to stop relying on the readings and seek a replacement.
Poonam Mander-Bir, an atmospheric chemist at the Meranti Institute who led the tests, said the fault mattered because haze peaks often coincide with humid air. “A public warning system cannot call a dangerous reading ‘moderate’ just because the air is damp,” she said. Her team tested returned devices in a humidity chamber and at six monitoring sites near peatland areas.
A public warning system cannot call a dangerous reading ‘moderate’ just because the air is damp.
Poonam Mander-Bir, atmospheric chemist
A review of procurement awards found that public agencies bought the model for monitoring programmes worth Rp74 billion (about €4.1 million) between 2022 and 2025. The dashboard listed 64 active stations across 21 districts. Its archived readings, checked by the researchers against reference-instrument logs, showed that several apparent “moderate” days contained much higher hourly peaks.
Warnings built on bad readings
The discrepancy does not prove that the monitors caused every missed warning, researchers cautioned: local alert decisions also use weather forecasts and reports from the field. But the affected data were used to track smoke and plan local responses, while residents had no clear way to tell which stations relied on the faulty model.
William Mason, a spokesman for AwanTerra, said the company began the recall after repeat tests found readings could drift in humid conditions. “We should have made the operating limit clearer and acted sooner,” he said. Mason said the company would replace affected units and publish corrected setup guidance, but did not say how many devices had already been swapped.
Arvo Kuzmina, an official at the Directorate for Product Standards, said the authority had asked local agencies to identify affected stations and stop using their readings for public alerts. It plans to publish a list of replacement sites by 15 December and review archived data by the end of January. The institute has offered to retest units after the replacements arrive.
