A Blockchain-based Predictive Maintenance Scheme for Smart Agriculture

Abstract

With technology invading all the sectors, the agriculture sector is not alien to technological developments and advancements. The use of IoT, robotics, sensors, artificial intelligence, etc., has resulted in the concept of Smart Agriculture, which aims to decrease human effort and increase the quantity and quality of the products. A lot of machinery is being used in smart agriculture and its maintenance is an important aspect for continued productivity and efficiency. Implementing a predictive maintenance scheme will reduce the high cost in terms of money and time and decrease unexpected faults and breakdowns. Motivated from the aforementioned discussion, in this paper, we propose a blockchain-based predictive maintenance scheme that utilizes the data collected by sensors to monitor and predict the condition of machines. The proposed scheme that uses K-Nearest Neighbours, Random Forest and LSTM models have been used and their performance has been compared based on accuracy, precision, recall, and F1-score. Results show that Random Forest achieved an F1-score of 89%, which is the best performance among all the models and can be effectively utilized for this purpose.