Progress in the last 20 years in airborne-, space- and drone-based imaging spectroscopy has advanced tremendously. These innovations have allowed improved large-scale monitoring of crop physiological processes with unprecedented detail. Successes have been obtained in the context of biotic and abiotic stress detection, particularly with new developments in sensor miniaturisation and physically-driven and artificial intelligence-driven modelling techniques. In 20 years, the spectral detail employed to detect stress has been exceptionally enhanced: cameras and technological imaging devices have moved from gathering data at the ‘hundreds of nanometers’ spectral scale down to the ‘sub-nanometer’ resolution, even reaching the Armstrong physical unit. Due to these rapid technological developments, the main focus has shifted recently, moving from a technology push in the last decade to the current algorithm-push to understand better the physiologcal interactions of crops undergoing biotic- and abiotic-induced stress. Xylella fastidiosa is currently the major transboundary plant pest, the number one threat for Australia, and the world’s most damaging pathogen in terms of socio-economic impact. As with several other pathogens under natural crop conditions, i.e. where the abiotic-induced variability due to water and nutrients co-exists with the pathogen-induced stress, its detection requires advanced remote sensing monitoring technology and algorithms to disentangle the biotic vs abiotic physiological interactions. These advanced methods use high-resolution hyperspectral and thermal imaging cameras onboard drones and piloted aircraft, demonstrating that uncoupling the biotic–abiotic spectral dynamics reduces the uncertainty in the disease detection, reaching accuracies over 90%. Although most currently operated drones are not carrying imaging spectrometers, efforts should be made to enable advanced remote sensing technology and algorithms with low-cost and easy to operate platforms for widespread hyperspectral technologies worldwide. These hyperspectral methods coupled with proper algorithms will advance the early detection of devastating pathogens, to reduce billions of losses worldwide.