An IoT solution for measuring bee pollination efficacy

In light of `saving the bees', it becomes necessary to learn as much as possible about bees and other pollinating insects. Data about flying patterns of bees, intensity of bees, wingbeat frequency of bees and activity of bees are all factors that can contribute to optimizing pollination by insects. Somehow, the pollination process is often overlooked in new internet of things- (IoT-) developments for agriculture. Although individual bees can already be tracked, there has yet to be found a way to actually map bee activity on a field. In this paper, a system is presented that enables fruit growers to detect, quantify and optimize bee activity, based on sensor technology. The system can predict and map the behavior of pollinating flying insects in an affordable, robust and location-based manner. Firstly, the state of the art on bee mortality and on IoT in agriculture is briefly explored. Secondly, a literature study is performed on detectable bee characteristics and sound is described as a good way of affordable detection. Thereafter, specifications for a practical IoT solution are provided. Lastly, for one of the most critical aspects, a machine learning algorithm for sound based detection of pollinating insects, is constructed and tested.