Successful handheld instruments are designed to give answers to non-scientist operators, and therefore their developers have put extensive resources into reliable identification algorithms, and qualitative and quantitative calibrations. LIBS builds on more than 60 years of experience in optical emission (arc-spark and inductively-coupled plasma) spectroscopy, while smartphone spectroscopy leverages existing colorimetric assays and their ‘chemistries’. Handheld near-infrared analyzers have similarly built on the decades of calibrations on laboratory instruments in food, feed and agriculture. The low-cost consumer instruments use a different model: crowd-sourcing for the company, and also crowd-sourcing for the data, with analysis in the cloud, not on the instrument itself.
This paper outlines the portable spectrometer field, and discusses databases, calibrations and algorithms, and also caveats on crowd-sourced data, especially for heterogeneous samples.