A UAV-Assisted Multirobot Network for Surveillance and Tracking Applications: Design and Implementation

The developments of AI in the security sector and related fields have made it possible to design and implement advanced robotics systems working with UAVs (or Drones) for applications such as surveillance, monitoring, and scanning for intrusions or breaches. However, such systems usually exhibit some challenges, difficulties and limitations in real-life implementation as these need to perform pre-mapping by scanning the area first and learning the paths, then encoding the resulting maps. This costs time and is complicated to implement. In this work, we design, implement, and test a system of drone and ground vehicles connected over a wireless communication network. The system performs different monitoring tasks paired with detailed analysis of regions of concern by the robots. The integration of a UAV in a form of a drone, gives the system an advantage of a moving bird-eye view of the area to be scanned. This advantage helps detects breaches that might be blocked as well as eliminating the blind spot limitations of regular CCTVs. When such a breach is detected, it locates itself and sends the location to the nearest ground vehicle which will in turn automatically navigate to that location for more insight into the type of breach or fault. With this system, we provide an efficient solution to the complex problem of localization and path planning while maintaining accuracy and fast response time. The developed system is capable of detecting intrusions and reporting breaches.