Answers are grounded in this paper and cite their sources.
Connected spirometers require reliable acquisition and transmission of turbine-generated events during respiratory maneuvers. This study evaluates event preservation, overload behavior, and browser-based reconstruction in the PHUKUY ESP32–Bluetooth Low Energy (BLE) subsystem. The evaluation comprised 196 runs: 30 controlled BISBC bench trials, 101 manual 3 L syringe strokes, 40 baseline synthetic trials, and 25 synthetic trials with online reconstruction. The BISBC low-flow profiles reached a maximum accepted event frequency of 226.65 Hz (FAST-profile mean PEF: 2.84 L/s), leaving calibration control points above 300 Hz unexercised by that bench. All 131 physical runs preserved firmware-observed events. In the syringe campaign, independent accounting and cumulative CRC-32 checks confirmed reception of all 24,720 observed and submitted events in 10,407 packets, without ring-buffer drops. Accepted inter-event frequencies reached 1865.67 Hz, whereas the maximum 100 ms window-average rate was 1380 events/s. Descriptive syringe volume reconstruction yielded 3.007 ± 0.031 L (CV: 1.045%) against a nominal 3 L stroke, without constituting independent calibration validation. Across all campaigns, all 2,190,639 submitted events were recovered, while 139,635 ring-buffer rejections occurred exclusively under synthetic overload. In baseline tests, five of five 10 s runs at each rate from 1500 to 2500 events/s were loss-free, with intermittent losses beginning at 3000 events/s. Online reconstruction agreed with post-capture replay using the same algorithm in all 25 trials. The loss-free 2500-events/s synthetic result is implementation-specific and does not establish universal BLE throughput or the maximum physical pulse-capture rate of the optical/GPIO front end.
No takes yet. Share an insight, caveat, or question.
Chavez et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: