The ASETA project (acronym for Adaptive Surveying and Early treatment of crops with a Team of Autonomous vehicles) is a multi-disciplinary project combining cooperating airborne and ground-based vehicles with advanced sensors and automated analysis to implement a smart treatment of weeds in agricultural fields. The purpose is to control and reduce the amount of herbicides, consumed energy and vehicle emissions in the weed detection and treatment process, thus reducing the environmental impact. The project addresses this issue through a closed loop cooperation among a team of unmanned aircraft system (UAS) and unmanned ground vehicles (UGV) with advanced vision sensors for 3D and multispectral imaging. This paper presents the scientific and technological challenges in the project, which include multivehicle estimation and guidance, heterogeneous multi-agent systems, task generation and allocation, remote sensing and 3D computer vision.