The dataset ranks 100 assignees by record count, and the top filer alone holds 244 of the 643 records in scope. The next four leading assignees bring the combined top-5 total to 329 records, or 51.2% of all records, before volume drops off sharply into a long tail of single- and few-filing entrants. This concentration means a small number of companies hold broad foundational positions while the rest of the field files narrower, more specific claims around them.

Filing peaked at 133 records in 2018 and has not returned to that level since. Using the most recent complete comparison, filings fell from 40 in 2021 to 18 in 2024, a -55% change over three years. Figures for 2025 and 2026 are still incomplete because publication typically lags filing by around 18 months, so they should not yet be read as evidence of continued decline.

Despite the search focus on particulate matter and gas sensing, the largest IPC class by share is G05B (control and regulating systems) at 42.8% of the 643 records, followed by H04L (digital transmission, 38.1%) and G06N (AI-based computing, 34.1%). Core sensing-related classes such as G01N (material analysis) and G01M (testing) sit lower, at 27.5% and 16.3% respectively. This indicates that recent patent activity is weighted toward how sensor data is processed, calibrated and transmitted, not just the physical sensing mechanism.

US20200393351A1, assigned to AMS International AG and published in December 2020, describes an integrated particulate matter sensor system built around a substrate, a cover defining a microfluidic flow path, and a fluid circulation device that moves air through an integrated sensor element. It is a specific architectural claim around fluid-path integration rather than a claim over particulate sensing broadly. Anyone designing a similarly integrated flow-path sensor module should review its claims closely, but the surrounding filing landscape shows plenty of room for different sensor-housing and flow-path architectures that do not read on this specific structure.

Class-level data shows thinner coverage in areas like cross-sensitivity compensation, low-cost calibration-reference modules and sensor-drift self-correction relative to the dense control and connectivity classes (G05B and H04L). Co-filing is also rare, with only 7 co-assignee pairs identified across the whole dataset, suggesting most companies work independently rather than pooling IP through joint ventures. Combined with a long tail of single-filing entrants outside the top 10 (which together hold 60.7% of records), this points to open ground in narrowly scoped sensing-hardware improvements rather than in the control layer, where the leading assignees are already dense.