Breadcrumb

Project to improve unmanned aerial vehicles

Published

Gerardo Schneider and Sebastián Uchitel

Autonomous cyber-physical systems (ACPS) consist of smart and collaborative computational entities, interacting with physical components to accomplish tasks in unpredictable environments. They are emerging from rapid advancements in research areas like data acquisition, machine learning, embedded systems and formal methods, and have countless areas of use.

Gerardo Schneider, University of Gothenburg, together with Sebastián Uchitel, University of Buenos Aires, have received a STINT Initiation grant for the project “Runtime Verification of Autonomous Systems”, in which they will attempt to program a UAV, an unmanned aerial vehicle, a special type of ACPS that can be used for instance for land recognition in rural areas. The system interacts tightly with physical components through sensors and activators, and needs to be able to accomplish complex tasks in unpredictable environments, independently or with minimal supervision from humans.

“We are interested in building a UAV on a fixed wing airplane based on an architecture using Discrete Event Control. We will build upon a declarative mission specification, for example search and rescue, follow or surveillance, of a controller that runs on the plane and interacts with lower software layers that takes high-level commands in an interactive and flexible manner” explain Sebastián and Gerardo.
The scientific aim of the project is to solve this problem allowing for dynamic mission planning with the aid of Runtime Verification, a technique for system monitoring and verification at runtime. From the educational point of view the objective is to promote dissemination of the knowledge to graduated students (in the form of PhD courses) as well as open presentations for a more general public.

The project was launched in August with a workshop and a mini-course at the University of Gothenburg with visiting lecturers from the University of Buenos Aires.