Sussex research student Allister Furey is keen to get his new project off the ground, as it could lead to a more effective way of harnessing wind energy – using kite-flying robots.
Allister is starting the second year of his doctorate in the Centre for Computational Neuroscience and Robotics (CCNR), working under the evolutionary robotics pioneer Dr Inman Harvey.
His research involves a collaboration with an Italian company, Kite Gen Research, and UK-based kite manufacturer Flexifoil to develop new generations of software to control kite wind turbines.
Kite Gen has developed the hardware for the kite wind generator – a small tower with limbs to which controllable arc-shaped giant kites are attached by cables. The kite movements are controlled by sensors and software – the robot. The aim is to fly the kites in a sideways figure-of-eight movement, for optimum power generation. The movement of the kites drives the turbine, creating electricity.
“Anyone who has flown a big kite and felt it pull will know the potential power there,” says Allister, who is also a keen kite-surfer. “And the amount of energy you get from flying the figure-of-eight pattern is massive.”
However, to be most effective, the technology must cope with extreme conditions, such as sudden gusts or lulls in wind.
Flying several kites in formation could also pose problems, without a highly sophisticated control system. Allister says: “If you have 12 kites flying together they could cross over and get tangled. What happens to each kite depends on both the wind and the steering movements.”
This is where Allister’s work with evolutionary robotics comes in. He will be developing a sensitive control system that can work with and adapt to wind conditions independently, applying biological principles of evolution and natural selection to create ‘artificial nervous systems’ for robots.
Allister says: “The conventional control method [for example an autopilot system in a plane] would involve fighting the turbulence, but in biology, things often work with turbulence. Tuna fish, for example, boost their speed by creating swirling movements of water with their body in a way almost impossible to design with conventional techniques.
At present, Allister is testing an evolved robot control system by subjecting it to a range of wind conditions in simulation. He is also working with CNNR tutor Bill Bigge to build a robot and test it by getting it to fly a stunt kite in real-world conditions.
Kite-driven generators could potentially produce energy for approximately half the cost of a coal-fired power station, while taking up 50-100 times less land than a typical wind farm.
Allister says: “To produce 5-10 megawatts of energy [enough to power over 5,000 homes] you would need a 500m2 kite, which is not far off the blade area of a typical offshore wind turbine. However, the kite system would be far cheaper to build, and less land is needed as the kites can be attached to cables, not towers, and can therefore be grouped together more closely.”
Kite wind turbines also have less visual impact than traditional wind farms and can utilise the more powerful and consistent winds at higher altitudes (88-1,000m) that traditional wind turbines cannot reach.