It’s a hot summer night and there’s another forest fire in Greece. This one has already burnt out 1000 hectares of a Mediterranean forest some 80 kms north of Athens.
“We can locate the worst of the blaze and also where the emergency services are. If they’re in the vicinity, we can decide how to direct them straight to the spot,” he continues.The flying robot is called a ‘microdrone’. Its sophisticated cameras allow it to see from above, where humans can´t always see – to the very heart of the forest fire. It can in this way give valuable information to the firefighters, helping them to better carry out their dangerous task. Chief Fire Officer of Central Greece, Nickos Floros, says the microdrone has already proven useful: “During the night, planes and helicopters are forbidden to fly over forest fires. But this robot can see where our firefighters can´t go, for example, on the other side of a burning slope, as is the case tonight.” “It can give us a quick glimpse of what is happening at the very heart of the fire, so we can decide what are the best options.” The microdrone is the result of a European research project aimed at providing a flying tool to help manage a crisis situation from above.
It weighs less than 1.5 kilograms; it can fly alone for up to around one hour; and it can carrry around 200 grams of material.Airrobot Managing Director Burkhard Wiggerich says: “It works with four motors, two that turn to the left and two that turn to the right. This makes it very stable in the air.” “It’s well equipped electronically, to make the job of flying or controlling it easier. At the same time, it positions itself using its GPS.” Detecting forest fires is just one of its possible uses. Developers think it could help manage other natural disasters like floods or help control urban riots.
But before all of these applications can become reality, there are some technical challenges to overcome.Microdrones Project Coordinator Christophe Leroux says: “Some of the failures of the GPS system on the drone have to be redressed. For example, when we’re too close to an obstacle, the drone can’t position itself and it begins to behave erratically and make mistakes. The danger here is that we could lose the machine, or damage it.” “In the future, what could be interesting is to give these machines more autonomy, or self-sufficiency, so that their operators can concentrate on what they have to do, and forget the technical problems and complexity,” Leroux continues. That is why computer scientists like German Chunrong Yuan from the University of Tubingen are developing what they call “environment sensing tools”. Sophisticated algorithmic calculations allow the drone to identify moving or stationary obstacles below – all by itself. Chunrong Yuan explains how they work on the microdrone: “The black points are static obstacles, and the red points show how the people are moving in the scene. This mechanism has been developed for the purpose of obstacle avoidance.” “During navigation it is more difficult to identify the moving obstacles – mainly because they are moving and the camera embedded in the drone is also moving. So we need to differentiate the two,” she continues. In Paris, at the French Atomic Energy Commission or CEA, in robotics headquarters, researchers are working on developing systems to increase the autonomy or self-sufficiency of the drone. The final aim is to develop a kind of ‘autopilot’ for the microdrones. CEA-LIST Robotic Engineer Laurent Eck gives an example of how this works:“We’re going to teach the drone a trajectory. To do this it utilises a camera that faces the floor to memorise the route. For the return, it has learnt its way.” “Now I don’t operate the drone any more. It automatically uses the information that was programmed in when it went in the opposite direction and returns automatically to its departure point.” “So the objective is to relieve the operator of the task of piloting the robot, so he doesn’t have to focus on the drone, but instead he can concentrate on his mission of observing what’s happening,” Eck continues. Scientists have also developed virtual reality scenarios that will allow the microdrone to better prepare its work. There is software that can simulate different flying conditions, including very windy weather, or difficult geographic surroundings- to see how the drone copes in real time. Thales European Project Manager Manuel Carballeda says: “Here we see what the drone sees with its camera, so we can find out, comparing it with the trajectory programmed in before, if the drone is going to be able to see the zones that interest us.” “This lets us check that what was pre-programmed is going to be really carried out by the drone, all in simulation mode.” Far from the simulations, forest fires are a deadly reality every summer in Greece and other Mediterranean countries. So firefighters and volunteers eagerly welcome any extra reliable assistance. Part-time Firefighter Vassilis Petrakis says: “I hope that these cameras above us will help us understand what we’re going to face near the fire, so we can be better prepared.” “Our job is really difficult and dangerous. During the day, we get burnt. During the night, we can break our legs. So any extra help is very welcome.”