Smart farming is becoming more cost effective. Advances in standards and protocols driven by interest in the internet of things are allowing for more choice in design. The increased availability enables for designs that are applicable for the most challenging of applications, that of subsistence agriculture in the developing world. The single most serious effect of climate change to date in the global south is that of water supply. The well defined wet and dry seasons are no longer well defined. Solar power together with microelectronics on boards and low power short range RF links can provide sensor fusion and actuator control at moderate cost. Near real time monitoring and evaluation is more the challenge. The concentration of the present work is on monitoring in an internet of things environment. Initial efforts to employ a convolutional neural network for data flow control is described.
Citation Fiehn, Heinz Boehmer, et al. “Smart agriculture system based on deep learning.” Proceedings of the 2nd International Conference on Smart Digital Environment. 2018.