The term “crowdsourcing” was coined around ten years ago from the words “crowd” and “outsourcing”. The basic idea is that acquiring data, collecting funds or decision making is spread over the whole community of people who actively participate in such activities. Although the concept of crowdsourcing is relatively new, there are numerous successful examples of its use. Big food producers let their consumers vote on the design of their next product, scientists get community’s help in analysis of astronomical objects and there is a huge number of start-ups and individuals whose products are funded through crowdfunding websites such as Kickstarter and Indiegogo.
More importantly, the concept of crowdsourcing opens up new opportunities for agriculture and BioSense Institute is currently working on such a platform for the Province of Vojvodina. Recently, affordable cameras that cover visible, NIR and thermal parts of the spectrum, as well as handheld Bluetooth spectrometers, have appeared on the market. They are not as precise as their expensive counterparts, but with the aid of advanced signal processing techniques developed at BioSense institute, we are able to extract vital information about the amount of water and nitrogen in plants, identify pests and propose adequate solutions. On top of that, information about farmer’s location, movement and type of crops planted serve as “ground truth”. “Ground truth” refers to real data collected on a particular location, which is compared to the output results of a model. The information about discrepancy is used for model’s fine tuning. In this case, it is used for calibration of our satellite crop classification system, thus increasing its accuracy.
BioSense aims to encourage farmers, agronomists, extension services and chemical vendors, to collect data on their fields mainly by taking photos with their smartphones, or more information from other sensors, if they are available. These images and measurements are then analysed by experts or fed to a larger system. In this way all of the stakeholders make the benefits for each other. Farmers get their plants analysed and receive recommendations on how to treat them, customised for location, soil condition, and the way their farm is managed. Agronomists and extension services get deeper insight into the plant condition and therefore can give better recommendations. On the other hand, scientists get the ground truth, while government institutions benefit from gathering large-scale statistics, crop-structure and yield predictions. This kind of statistics is of an exceptional value for strategic planning, price formation and determining the amount of subsidies.
Mobile-phone app developed at BioSense allows farmers to feed information directly into the system by simply taking photos of their crops. The crowdsourcing system is currently being upgraded with miniature thermal camera and spectrometer (also shown), to provide much deeper insight into the status of crops.