Tutorial - Gabriele Costante

Autonomous robots have established as a ground-breaking technology for precision farming (PA), particularly for resource optimization and labor savings. One of the most important aspects of PA that benefits from robotic solutions is certainly crop monitoring. Robotic platforms such as All-Terrain Vehicles (ATVs) and Unmanned Aerial Vehicles (UAVs) can now be equipped with high throughput embedded computers (also with GPUs) that open further possibilities for automatic inspection and, more generally, for smart agriculture. To close the gap with human capabilities, robots must use their perception systems to achieve three main goals: navigate the operational scenario, collect information, and process it. Unlike urban or indoor scenarios, agricultural settings offer specific challenges that must be considered when developing an autonomous monitoring platform. These include the presence of rough terrain, variable operating conditions (e.g., weather, lighting), and repetitive and non-discriminative objects that make localization and navigation more difficult. In this tutorial, the characteristics of a robotic platform for crop monitoring will be described, highlighting the hardware specifications, the sensors required to collect information about the surrounding environment, and the algorithms needed to enable the robot's autonomy and allow the processing of collected data. Different monitoring applications will be presented, including normalized vegetation difference index (NDVI) computation for prescription maps and olive fly infestation detection. As a case study, a ground robotic platform, developed as part of a research project, will be considered and presented after the tutorial session, showing examples of data acquisition sessions.

Gabriele Costante received the Master degree in 2012 and the Ph.D. degree in information engineering in 2016 from the Department of Engineering, University of Perugia, Italy. He is currently a Senior Researcher (RTD-B) with the Intelligent Systems, Automation and Robotics Laboratory (ISARLab) at the Department of Engineering of the University of Perugia. In 2014 he has participated as a visiting researcher in the activities of the Robotics and Perception Research Group of the University of Zurich (Zurich, Switzerland). He has co-authored more than 40 scientific papers in international journals and conferences. He is currently member of the Academic Board of the Ph.D. in Industrial and Information Engineering at the Department of Engineering, University of Perugia and lecturer of the courses “Computer Vision and Robot Perception” and “Machine Learning and Data Analysis” within the master degree in Informatics and Robotics Engineering. He has participated and is currently active, also with management responsibilities, in several national and international projects with universities and companies. He is also involved in different editorial activities and scientific committees and is co-founder of the academic spin-off Red Lynx Robotics s.r.l.. His research interests include artificial intelligence, robotics, computer vision, and machine learning with applications in several contexts including healthcare, agriculture, logistic and industry.