Supervision and monitoring of agricultural autonomous robots: from design to implementation

The integration of robots and IoT leads to the concepts of Internet of Robotic Things (IoRT). Agroecology aims to develop cultural practices leading to environmentally friendly farming production through crop diversification and biodiversity. IoT and drones could provide stakeholders with more precise, complete and innovative data allowing to define new agroecology practices, and robots perform repetitive and accurate farming operations over a long time period with a low impact on the environment. IoRT and agroecology lead to the concept of Agroecology 4.0, such as one particular field of Industry 4.0. Agroecology 4.0 solutions need a system for monitoring autonomous robots, which is not an easy task as it requires expertises in several fields: agronomy, communication networks, Big Data, IoT and autonomous vehicles.
The conceptual design of information systems is mandatory in several application domains. Motivated by the lack of a conceptual model for IoRT data, we propose a UML profile taking into account different kinds of data (e.g., sensors, stream, or transactional) and non-functional Requirements, and we present it in the context of the autonomous agricultural robots supervision application. We also provide the details of a new architecture for this agricultural task detailing the different data management layers. A focus is done on the Multi-model Data Warehouse to store and analyze these IoRT data. Multi-modelDBMSs (MMDBMSs) have been recently introduced to store and seamlessly queryheterogeneous data (structured, semi-structured, graph-based, etc.) in their native form, aimed at effectively preserving their variety. Unfortunately, when it comes to analyzing these data,traditional data warehouses (DWs) and OLAP systems fall short because they rely on relational DBMSs for storage and querying, thus constraining data variety into the rigidity of a structured, fixed schema. Therefore we propose the usage of an MMDBMS to store multidimensional data for OLAP analyses. A multi-modelDW would store each of its elements according to its native model; among the benefits we envision for this solution, that of bridging the architectural gap between data lakes and DWs, that of reducing the cost for ETL, and that of ensuring better flexibility, extensibility, and evolvability thanks to the combined use of structured and schemaless data.


Sandro Bimonte is Reseach Director at French National Research Institute for Agriculture, Food and the Environment (France), and more exactly he is member of the TSCF laboratory. He received his PhD from INSA-Lyon, France (2004–2007). He is an editorial review board member of International Journal of Data Warehousing and Mining, International Journal of Decision Support System Technology, and international conferences such as ER, DOLAP, etc. He has published more than 100 papers in refereed journals and international conferences. His research activities concern spatial data warehouses and spatial OLAP, visual languages, geographic information systems, spatio-temporal databases, geovisualisation, Big Data, and IoT. He joined and coordinated several research projects (such as, Superob, BEYOND) on the above areas.