AgriPrediction: A proactive internet of things model to anticipate problems and improve production in agricultural crops

•AgriPrediction combines LoRa technology, time series and the ARIMA prediction model.•The idea is to anticipate crop dysfunctions proactively, notifying the farmer for remedy actions.•We have a collection of time series, one for each sensor, providing a higher precision to the farmer.•In an arugula cultivation, gains were obtained in terms to leaf development and weight.•Our design support system is financially viable, costing close to $250 (USD currency). One of the significant challenges for the future is to guarantee food for all inhabitants of the planet. One of the alternatives for this issue consists in increasing the production, but to accomplish this, it is necessary that innovative options be applied to enhance the soil capacity and the protection of environmental resources. In this context, Internet of Things (IoT) is gaining more and more attention, with a lot of alternatives to aid farmers with smart sensors and visualization systems. However, the state-of-the-art still presents no other options of IoT applications in the rural environment that assist the agricultural producer in the decision making about when to act, or to anticipate problems, in the crops. This article presents a model named AgriPrediction, which combines a short and medium wireless network range system with a prediction engine to anticipate potential crop dysfunctions proactively, so notifying the farmer for remedial actions as soon as possible. To achieve this, AgriPrediction presents a framework whose components are based on both LoRa IoT technology and ARIMA prediction model. Our results demonstrated the feasibility of using LoRa in rural places, besides providing the advantages of having a prediction system to observe troubles related to soil humidity and temperature. In particular, when using AgriPrediction in arugula cultivation, gains of 17.94% were obtained concerning leaf development and 14.29% terms of weight in comparison with a standard cultivation procedure.