Students create APM-powered swarming drones with human-spotting cameras

From Engineering and Technology Magazine, a short piece on a cool project using APM 2.6 and a modified Skywalker X8:

A group of Loughborough University students have devised a ‘swarm’ system using smart unmanned aerial vehicles (UAVs) that could be the future for mountain search and rescue.

The 13 undergraduates were challenged to design and develop a UAV capable of finding a person in a search and rescue mission for a group project as part of their final year aeronautical MEng degree.

The swarm system uses up to ten UAVs operating together, flying at about 50-55 mph, to search an area of 12 square miles using GPS. Each has an infrared camera and they can ‘talk’ to each other. So as long as one is in range of base they can all communicate with the rescue team.

The students developed an image processing code that enables the cameras to detect human beings. However, due to the time constraints of a seven-month project, the images can only be downloaded and processed after the UAVs have returned to base.

The next stage is to do this in real time and the lecturers want a group in the next academic year to continue the work, developing a system which is potentially cheaper than using helicopters and quicker than rescuers on quad bikes.

Here's a bit more from the video description:

Undergraduates of Loughborough University have to do a Group Design Project as part of their final year Aeronautical MEng degree. This year, with the help of Simon Howroyd a PhD student at the University, one of the groups were challenged to design and develop a UAV capable of searching for a person, in a Search and Rescue (SAR) mission. The team known as "Overwatch" have heavily modified a Skywalker X8 and used APM2.6 to create a stable camera platform capable of flying search missions.The project goes beyond the typical hobbyist modelling approach and features heavy theoretical design and analysis including Computational Fluid Dynamics to assess the aerodynamics, backed up by wind tunnel testing. They have also done structural modelling to ensure a stiff structure in unforgiving weather conditions such as storms in mountainous areas which are hard to reach by rescuers en foot.It is relatively short project, only spanning around 7 months, with all the undergraduates having no prior knowledge of UAVs or RC aircraft, so the learning curve has been steep! Factor this alongside working with unusual materials in this field such as Styrofoam and expanded polyolefin (EPO) and you have a great project to vastly improve the employability of the students in industry after graduation.May 2014 sees the project come to a conclusion with a presentation and report due. Of particular interest the output of the system is a GPS location and photo of the person to be rescued. This is achieved by utilising a standard search pattern and onboard image processing of camera frames where Matlab code (recompiled as C code and uploaded onto a BeagleBone) makes decisions as to the probability of the subject being found in each frame.

Views: 2056

Comment by Julien Dubois on May 31, 2014 at 2:40am

Very interesting and usefull project!

Could you please develop that point?

Each has an infrared camera and they can ‘talk’ to each other. So as long as one is in range of base they can all communicate with the rescue team.

If APMs are communicating together, I would be interested watching their code!

Comment by John Githens on May 31, 2014 at 8:43am

Link to earlier, related DIYD blog post.


You need to be a member of DIY Drones to add comments!

Join DIY Drones


Season Two of the Trust Time Trial (T3) Contest 
A list of all T3 contests is here. The current round, the Vertical Horizontal one, is here

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service