Team/Robot name: Team: Team Flash Robot: Flash Introduction/Objective: A fully autonomous line follower which…
Aerial search and rescue (SAR) operations rescue more than 1700 people annually in the UK.Some of these cases are handled by voluntary mountain rescue teams. However, most of them require specialised training and equipment, thus the involvement of RAF in the UK or a government agency in other countries. In both situations, the work could be facilitated by the use of drones. Our solution is to develop a low-cost, quick-launch drone with surveying capabilities – this would speed up the work of the voluntary teams and reduce the cost of operating a helicopter for many hours. Here we emphasise on localisation – the most important aspect of a SAR operation.
There are many types of SAR operations:
These operations come at astronomical costs to emergency services worldwide. Aerial search has the advantage over ground based searches by covering a wider area in a shorter amount of time however the high costs, safety and training requirements of this limits its usage during SAR operations. The training requirements for a search and rescue pilot is often both a very expensive and time consuming commitment.
SAR, mainly Urban SAR can be closely related to disaster relief in urban areas, i.e. a flood or earthquake. A system which could monitor the area from above and provide realtime images/footage of the area would be useful to provide information to rescuers. A solution to this problem should attempt to reduce the cost of equipment, simplify the systems implementation and reduce the complexity of operation.
Our solution is to design simpler low cost alternative to traditional aerial SAR for area surveying. This is in the form of a semi-autonomous quadcopter drone. Existing drones on the market cost around £10’000 our aim is to produce the system for under £1000. The main features of our drone would be:
We would achieve this by designing our own lightweight(<3.5kg) quadcopter drone, using a commercially available frame and low cost components. The system will have a GPS module to help locate a person’s phone and will communicate with a PC/smartphone to relay information to the user while also having the ability to be controlled from a transmitter remote. The quadcopter aspect of this solution will be controlled by an open source Arduino-based flight controller (FC) called MultiWii 32 which will communicate to an electronic speed controller and control actions of the drone. An onboard Raspberry Pi (RPi) will be able to talk to the FC via an SPI bus, this would allow the Raspberry Pi to aid in locating a person, based on their phone GPS, and tell the drone where to go. The Raspberry Pi will communicate via a cellular link to the operator’s computer/smartphone to provide real-time images and location data, it has acquired from the onboard camera and GPS module. It should be noted that the transmitter controller remote by-passes the RPi control of the drone – this is a safety feature, should an issue arise with the users’ PC, due to either a local fault on their hardware, or if the cellular link is interrupted, and would allow the user to regain control of the drone directly to the FC.
Charlie Owens, Aleksandar Angelov, Joan Kangro, Sergio Martin Yebra, Cristian Chirion and Gabor Borics-Kurti