Comp 4 Drones

Automated Drones & Hyperspectral Technology for Construction sites

Within the COMP4Drones ECSEL JU project, Airobot demonstrated the added value of using a fully automated drone with hyperspectral technology on a construction site. During the project, together with IMEC Belgium, we demonstrated that hyperspectral technology is capable to distinguish between different types of soil. This technology can be carried by a fully automated drone to allow repetitive data collection. All communication is done via secure connections and cloud servers to guarantee quality and security.



The data is collected by a special version of the Airobot Mapper, which is capable to perform fully automated BVLOS flights.

The payload, designed by IMEC BE, is fully integrated with the drone. The cameras can be configured in real time in the pilot user interface, so no special links are needed.
After the flight, the data is easily downloaded from the payload via FTP and can be processed by the IMEC post-processing chain to generate a hyperspectral cube.

The AiroCollect cloud software has been modified to display the different bands of the hyperspectral cube. Users can combine this hyperspectral data easily with traditional RGB data and perform basic measurements. (e.g. soil volumes)


  • GNSS Jamming & Spoofing test
  • Evaluation of 4G network at different altitudes
  • Normal flight between 2 points
  • Aborted Take-Off during take-off
  • Aborted take-off within the security zone
  • Aborted take-off outside of the security zone
  • RC link loss
  • Manual take over
  • Companion computer loss
  • Abort landing
  • RTK landing on pre-determined landing position
  • Automated take-off and landing in box
Safe, Remotely Operated, Flight!

AIROBOT used the AiroCore to turn a standard drone platform into an automated BVLOS drone, controlled via 4G. After power on, the drone automatically creates a secure link via 4G to a server hosted by Airobot. Only an operator with the right credentials can get access to the drone via the same server.
Via a secure remote IP connection, the operator gets access to the onboard AiroFly software. In the AiroFly software, the operator can plan a flight (e.g. a mapping flight) and monitor the drone during the flight. The drone features several cameras (front, aft, top & bottom) to give the operator a complete view of the situation around the drone.
The drone features 4 camera’s (top, bottom, front, back) to give the remote pilot full view of the surroundings. A downwards facing lidar is available to maintain a certain ground distance and to assist during landing.
Key systems like GNSS, Compass and IMU have been duplicated for redundancy purposes. Systems from multiple manufacturers are being provided by the autopilot to increase the overall reliability. 
Next to the autopilot, the drone features 2 companion computers: one for BVLOS flight management and one for payload management, autoland, image processing etc.
On the AIROBOT Mapper, it is possible to have a direct secure IP link from a custom application to the payload to support custom interfaces.
Within the Comp4Drones project all tests were performed to demonstrate that the data collection can be done in a remotely operated BVLOS flight.


The drone is operated via the AiroFly application, which can be remotely accessed using a browser.
Within the AiroFly interface, the remote operator can plan the flight, view the drone’s surroundings, monitor all parameters and send override commands. Orthomosaic images generated with AiroCollect can be used as a layer in the screen to plan the flights.
A ‘Master Caution – Master Warning’ system has been implemented, similar to standard aviation. The system continuously monitors all hardware components and notifies the operator in case of a failure. A ‘Master Caution’ error, e.g.,  low battery, requires immediate action while a ‘Master Warning’ (e.g. failure of secondary GNSS) allows to continue the flight.

Safety Protocols have been implemented to support the required SORA analysis that is required for BVLOS (beyond visual line of sight) or EVLOS (extended visual line of sight)flights. Safety protocols haven been implemented for situation such as aborted take-off, aborted landing, infringement of airspace, loss of 2,4GHz back-up control link, loss of companion computer and low battery.
Also support for perimeters has been added to AiroFly (take off area, landing zone, alternate landing zone, 1:1 ground safety buffer.

Through the AiroFly interface it is possible to set the camera parameters (frame rate & integration time) and start the recording.
Also the output of both hyperspectral cameras can be monitored in real-time.  
With the data collected, it is possible to generate both standard RGB images and hyperspectral images.

The AiroCore is a single printed circuit board that integrates the autopilot, real-time flight controllers, a payload management computer, communication technology, an industrial multi-constellation GNSS receiver and redundant power supply. It can turn any multirotor, fixed wing or VTOL (Vertical Take-Off and Landing) drone into a ‘flying robot’.

  • It turns the drone into a flying robot so that it flies automatically with virtually no human intervention.
  • It captures and georeferences data in a standardized way, interfacing correctly with the GNSS receiver to achieve this.
  • It allows to use custom software to support customers’ own payloads and implement proprietary business logic.
  • It reduces wiring and interface cables traditionally used to connect all sensors and electronics inside a drone, and as such, it increases the reliability and manufacturability of the solution.

The AiroCore focuses on Beyond Visual line of sight (BVLOS) flying, via a 4G/5G/LTE network or local log range WIFI link.  The professional multi-constellation GNSS receiver is crucial to offer a more robust and reliable solution. 

Hyperspectral Payload and Pre-Processing (IMEC)

The IMEC UAV hyperspectral payload spans visible and NIR spectral ranges (450-970nm) and offers measurements in 32 spectral bands. The sensors developed by IMEC use a unique integrated hyperspectral filter/imager technology, where the spectral filters are monolithically deposited/integrated on top of CMOS image sensors at wafer level.
The pre-processing consists of four different sub-modules. First, a demosaicking algorithm is developed to estimate a multispectral image with full spatial-spectral definition. Second, a hyperspectral image cube is created. Third, several corrections are carried out: radiometric, non-uniformity, reflectance, spectral (varying lighting conditions) and corrections coping with degradations due to vibrations.


The AiroCollect is a scalable cloud-based software to process, store and share data collected with drones. Today, users can upload RGB images, which are used to automatically generate pointclouds and orthomosaics in which they can perform measurements (volumes, surfaces, distances…) online.
Within the Comp4Drones project, AiroCollect has been updated to also load and display the different hyperspectral bands.

COMP4DRONES is an ECSEL JU project coordinated by Indra that brings together a consortium of 48 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. https://www.comp4drones.eu/


Scroll to Top


Please fill in the information with * below
Press demo video to have direct access