Application of extreme machine learning for smart agricultural robots to reduce manoeuvering adaptability errors

The agricultural robots limit complicated manoeuvring jobs, reducing the need for human intervention. Automation and control of these robots are performed by observing and practising the crop type, regulations, and external weather conditions. This article proposes and discusses a Manoeuvering Adaptable Task Processing Model (MATPM). This model is designed to improve the adaptability and precision of agricultural robots coinciding with farming and climatic conditions. For this purpose, extreme machine learning is employed to learn, align, and respond to external conditions to improve precise manoeuvring. The external impacting features and their adaptability to the crop and climatic conditions are valued using the learning process. In this process, maximum adaptability is considered to improve precision agriculture. The tasks are then classified based on adaptability and executed sequentially. If an unpredictable impacting factor hinders a task, then the adaptability is paused from training and it improves the chances of training by using different completed and paused tasks from the previous manoeuvring processes. Therefore, precision in smart farming and robot system adaptability is ensured with fewer adaptability errors. From the comparative assessment, the proposed model improves adaptability by 8.71 % and precision by 11.44 %, with 8.96 % less adaptability error for various tasks/ day.