Artificial Intelligence-Driven Autonomous Robot for Precision Agriculture

Abstract

In the recent years, robotic systems became more advanced and more accessible. This has led to their slow, but stable integration and use in different processes and applications, including in the agricultural domain. Nowadays, agricultural robots are developed with the aim to replace the human labour in the otherwise exhausting, time-consuming or dangerous activities. Agricultural robotic systems provide many advantages, which can differ based on the type of the robot and its sensors, actuators and communication systems. This paper presents the design, the construction process, the main characteristics and the evaluation of a prototype of a small-scale agricultural robot that can be used for some of the simplest activities in agricultural enterprises. The robot is designed as an end-user autonomous mobile system, which is capable of self-localization and can map or inspect a specific farming area. The decision-making capabilities of the robot are based on artificial intelligence (AI) algorithms, which allow it to perform specific actions in accordance to the situation and the surrounding environment. The presented prototype is in its early development and evaluation stages and the paper concludes with discussions on the possible further improvements of the platform.

Keywords

• computer vision algorithms
• robotics
• simultaneous localization and mapping
• visual navigation
• visual odometry
• smart farming