Towards an Architecture for Agricultural Autonomous Robots’ Scheduling

Smart farming and IoT technologies open up a new research agenda, which relates to different inter-related scopes within a Farm Management Information System, such as robots’ programming, tasks’ scheduling, sensor data capture, management and processing at different layers of the IoT ecosystem. Many research works address these topics, but to the best of our knowledge none has contributed with a fully-featured architecture design of monitoring and scheduling of autonomous agricultural robots. In this paper, we propose the skeleton of architecture for such kind of IoT systems, called LambdAgrIoT. It is designed to support big data and different types of workload (real-time, near real-time, analytic, and transactional). We present the main features of each layer, and the implementation details and its deployment of the Data Source and Speed layers in a real environment. The paper also discusses the open issues related to the other layers and the deployment of the overall architecture at large scale.