LGMay 4, 2022

Modelling calibration uncertainty in networks of environmental sensors

arXiv:2205.01988v27 citationsh-index: 19
Originality Incremental advance
AI Analysis

This addresses calibration uncertainty, a key barrier for deploying low-cost sensor networks and enabling citizen-science research, though it appears incremental as an extension of existing calibration methods.

The paper tackles the problem of poor accuracy and drift in low-cost environmental sensor networks by proposing a variational approach to model calibration with uncertainty, achieving better performance than state-of-the-art multi-hop calibration on synthetic and real air pollution data.

Networks of low-cost sensors are becoming ubiquitous, but often suffer from poor accuracies and drift. Regular colocation with reference sensors allows recalibration but is complicated and expensive. Alternatively the calibration can be transferred using low-cost, mobile sensors. However inferring the calibration (with uncertainty) becomes difficult. We propose a variational approach to model the calibration across the network. We demonstrate the approach on synthetic and real air pollution data, and find it can perform better than the state of the art (multi-hop calibration). We extend it to categorical data produced by citizen-scientist labelling. In Summary: The method achieves uncertainty-quantified calibration, which has been one of the barriers to low-cost sensor deployment and citizen-science research.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes