Low-cost and autonomous hyperspectral sensor node for precision agriculture methods

Precision agriculture (PA) is a complete farm management approach that uses and exploits different sources of information at different scales (spatial, temporal and spectral) provided by varied technology (sensors, satellites, UAVs, robots) for improved resource use efficiency, productivity, quality, profitability and sustainability of agricultural production.
One of the common methods used in PA is spectroscopy. It relies in the analysis of the electromagnetic spectrum in the visible and near infrared ranges. Some of the applications are quality check of fruits, meat, and fish, grading and sorting at-harvest applications for fruits and vegetables, soil characterization, health of vegetation for monitoring diseases, and many others.
Traditionally, the instruments used in spectroscopy are costly and oriented to laboratory applications. Also exist field instruments but the limited temporal resolution impacts in the overall analysis of the applications.
In this work, we develop a low-cost and autonomous hyperspectral sensor node that is able to collect daily spectral signatures, for the quasi-real time analysis in an apple orchard affected by apple proliferation disease. The system is based in Raspberry Pi and Arduino platforms using a low-cost spectral sensor.