Smart Sensors

Abstract

Calibration costs still limit the mass deployment of smart systems based on low-cost chemical sensors. Recent cost-efficient calibration strategies skip the acquisition of a full set of calibration samples when building new models. For example, calibration transfer and calibration update take advantage of models previously built using a reference system. Wireless sensor network calibration can also leverage the effects of intra-node correlations. In this chapter, I review such strategies together with public datasets that are valuable for benchmarking calibration methodologies for low-cost chemical sensors.