Comp4Drones (EN)

COMP4DRONES is an ECSEL JU project coordinated by Indra that brings together a consortium of 49 partners with the aim of providing a framework of key enabling technologies for safe and autonomous drones. It brings to bear a holistically designed ecosystem from application to electronic components, realized as a tightly integrated multi-vendor and compositional UAV embedded architecture solution and a tool chain complementing the compositional architecture principles.

The project will mainly focus on the following objectives:

• Digitization of the state of the construction process of an infrastructure.
• Definition of the state of the construction process of an underground infrastructure.

In order to achieve the objectives, real drone flights, LIDAR point captures, RGB images and the use of simulators to extend the sensed data will be carried out, allowing us to generate autonomous detection systems (based on Deep Learning techniques with CNN networks) to identify elements related to the geometry and characteristics of the site terrain, the flat surfaces of the structures, the recognition of fixed elements and alignments and once detected, to create a database with objects of the same type.

In the same way, and applying the same Deep Learning techniques to achieve the second objective, a drone will be flown inside the underground infrastructure, to carry out the detection of any obstacle that interferes inside the tunnel. In this way, the efficiency of the planning activities, exploration, measurements of the underground environment, tunnel cataloguing, generation of high precision models, etc. will be helped.