The autonomous and coordinated navigation of the UAV swarm is a research topic that is based on the small drones that are equipped with GNSS devices and perception systems. The embedded processing system is a Jetson AGX Xavier, which has a 512-core Volta GPU with Tensor Cores and 8-core ARM v8.2 64-bit CPU, 8MB L2 + 4MB L3, and provides a full-featured development platform designed to get you up and running quickly. So, a drone swarm is composed of small drones that allow the development and testing of autonomous and cooperative navigation algorithms. The development of smart architectures also includes a set of complementary methods, such as the establishment of different control layers, the algorithms for obstacle detection and avoidance, and the methods for autonomous decision making. The possibility of being able to deploy in the same area a fleet of UAVs, that are capable of working in a coordinated way, is a smart solution to solve different collaborative tasks in an efficient manner in many different fields, like agriculture or forest fires, where the drone swarm can undertake, simultaneously, different tasks such as monitoring a certain region of interest, monitoring of land equipment deployed in the area, capturing and processing essential information, among others. So, the advanced research on drone swarm allows the team of drones to work together and the possibility of navigating and undertaking work autonomously, without the need for human supervision deployed in the affected area, thus reducing the exposure of people to dangerous environments, and reducing the possibility of personal harm.
Intelligent Systems Lab collaborates with Drone Hopper company to develop heavy payload multirotor platforms to be used as firefighting quick-response UAVs, capable of transporting and realising water and additives, at a low altitude, over the fire. The platform consists of a hexacopter multirotor powered by 6 propellers, which enables quick and efficient response to common uncontrolled fires and support each firefighting brigade to increase flexibility and overall performance. The firefighting UAV is based on a novel methodology to reach the objectives of an autonomous firefighting drone in forest and urban environments, where the use of perception systems, such as cameras and multi-plane lasers, and satellite navigation devices allow to extract accurate environment data and analyse them in order to extract information and knowledge of interest to extinguish the fire of the environment. The data fusion of perception systems and positioning by satellite navigation is a mandatory issue and allows to indicate the physical presence of aerial obstacles within the field of view of the autonomous drone by identifying visual attributes and classifying the overall information to be translated to knowledge. So, vision-based systems, laser scanners, inertial measurement units and multi-constellation GNSS devices are capable to enrich and improve the data fusion structures into low-high level control architectures to activate actuators in real-time to accomplish complex and safe active manoeuvres of the firefighting UAV. That is, the embedded architecture focuses on the injection of accurate perception data and self-flying knowledge to the automatic control embedded device of the high-capabilities UAV with accurate decision-making to estimate the complex behaviour of autonomous manoeuvres in firefighting environments.
Intelligent Systems Lab collaborates with Vías y Construcciones and Tecsa railway companies to design and develop a tethered drone based on a secure physical connection system using power and data cable connected to autonomous railway prototype that provides, uninterrupted power supply to the autonomous aerial vehicle and high-speed data connection for sending aerial images and for the comprehensive inspection of aerial elements and complete digitization of the rail track in real-time during a full inspection mission of the aerial railway infrastructure. The safe cabling system keeps the UAV within the safety gauge of the railway catenary. The design and development of this system is feasible using advanced control techniques for the hover and navigation of aerial devices connected by cable. The main feature of this captive drone is the guarantee of railway safety because it is tethered by a cable and its maximum extension is 2.5 meters, therefore, this limitation in the extension of the cable eliminates the risk of contact with the catenary cables or the railway infrastructure. In addition, it has continuous power so it can have several hours of service without having to stop every 15 minutes to recharge or change the batteries.