Technology to avoid mid-air collisions will be vital when we've all got our own helicopters

WHO wouldn't like to have their own personal flying machine tucked away in the garage? It's a nice thought, if perhaps a little far-fetched.

Be that as it may, the European Commission is taking the prospect seriously. And it is already worried about how squadrons of non-expert pilots are going to cope with this three-dimensional freedom. Will air hogs cause mid-air collisions?

WHO wouldn't like to have their own personal flying machine tucked away in the garage? It's a nice thought, if perhaps a little far-fetched.

Be that as it may, the European Commission is taking the prospect seriously. And it is already worried about how squadrons of non-expert pilots are going to cope with this three-dimensional freedom. Will air hogs cause mid-air collisions?

That's why the EC has kicked off a €4.3 million research project, called MyCopter, that aims to ensure "personal aerial vehicles" (PAVs) can fly automatically in neat, well-spaced "swarms" by sensing the other vehicles around them.

"It is now a question of when we'll get personal aerial vehicles, not if we'll get them," claims Heinrich Bülthoff of the Max Planck Institute for Biological Cybernetics in Tübingen, Germany, who heads the project.

Why does he think the idea of PAVs is not quite so fantastic? Witness Terrafugia, he says, a spin-off from the Massachusetts Institute of Technology, based in Woburn. The firm hopes next year to launch the Transition, a lightweight propeller-powered aircraft that can land on a runway, fold its wings and be driven away like a car - although not everything has gone to plan (see "Flying cars: a chequered history"). What's more, technologies for lofting and propelling small aircraft are maturing fast, Bülthoff says, thanks to recent developments in helicopter drones.

Despite Terrafugia's progress, MyCopter is only focusing on helicopter-style PAVs because they won't need a runway: you can take off from a parking space. "We are looking at vertical take-off and landing because we don't want to use airports for air commuting. There's a problem, however: we don't expect the average car driver to know how to fly a helicopter."

To begin to address this, MyCopter project members - including researchers at the University of Liverpool in the UK and the Swiss Federal Institute of Technology in Lausanne (EPFL) - aim to establish rules for autonomous flight, including sensing, flocking, control and simulation, and to develop simple user interfaces for pilots.

Accurate sensing will be at the heart of it all. At EPFL, Dario Floreano's team is investigating what mix of anti-collision sensors PAVs might need. He envisages each craft pinging radio signals to all the others and measuring the signals' return time to keep a safe distance. "It will work even when GPS is not available," Floreano says, adding that such a system is now being tested on a swarm of 10 four-rotor robots, as part of the EC-funded Swarmanoid project.

Floreano's team has already developed image sensors with a wraparound view like that of an insect's compound eye. If visibility is poor, acoustic sensors may do the trick. "You can work out the speed and direction of approaching aircraft from the sound they make," Floreano says. Radar and laser-based sensors (lidar) will also be tested. The EPFL team hopes to build a demonstration sensor array by 2014.

PAVs might not be entirely autonomous, says Bülthoff. You should be able to override them, just as drivers can ignore satnav directions in a car, he says.

Fortunately, MyCopter will have other applications before we eventually take delivery of our first personal helicopter: the flocking algorithms developed will make it easier for drones to fly in civilian airspace, and simpler user interfaces could make it easier to fly them. As for replacing cars, the technology's there, says Bülthoff. "Making it affordable is another question."