For characterized climacteric fruit, its sensitivity to ethylene can be weakened by reducing the concentration of exogenous ethylene, so as to postpone the emergence of endogenous ethylene climacteric peak and respiratory climacteric peak, which plays a role in postponing the senescence of fruits to some extent. Firstly, the MOX gas sensor was adopted to build an ethylene concentration sensor with high precision;secondly, it carried out single and multiple gas sources localization experiments based on Kernel DM+V algorithm to form a SLAM-GDM map to locate the gas source and visualize the concentration distribution of ethylene in an unknown environment;finally, it performed the degradation experiment by using ethylene degradation robots. The experiment findings indicated that the built gas distribution model can predict the gas source localization. The average error of the SLAM-GDM map performance in predicting the single gas source localization was 0.78m with more than 95% of accuracy. The multiple gas sources localization prediction experiment showed that the analysis of variance presented high prediction accuracy. The average error of the gas source prediction localization was less than 048m with less than 5% of relative error, and the gas degradation robot can accurately find the gas source position according to the ethylene gas distribution mapping, and conduct ethylene degradation at a high rate. The ethylene gas-source was positioned and degraded ethylene through mobile robots, to extend the storage and shelf life of fruits and it had an application prospect in the field of fruit preservation.