Fruit flies and robots might not seem to have much in common, until a biologist at the University of Washington, filmed, documented and studied fruit fly flight patterns. Biologist Michael Dickinson is obsessed with fruit fly flight, and has spent years examining insect flight patterns. This information, when properly applied, can be used to build robots the size of flies for a multitude of activities. Search and rescue missions where people have no immediate access to search for trapped individuals, for example. Dickinson has a miniature arena in a laboratory in Seattle, three high-speed cameras, and a 7500 frame speed to film every turn, dip and roll of his fruit flies.
His current work concerns the physics of how flies escape different threats with a specific sequence of rapid wing beats and sharp turns. The magazine Science recently published his paper on the physics of the fruit flies‘ maneuvers. Engineers may soon be able to apply this research in building smaller and smarter flying robots.
Until Dickinson filmed the actions, it was unknown if fruit flies use visual cues to generate elusive maneuvers, and initiate rapid, banked turns, to avoid a threat. Dickinson spoke enthusiastically about the confirmed results, and exclaimed they do indeed use visual cues to avoid disaster.
By examining 3,566 wing beats from 92 assorted fruit flies, Dickinson and his team discovered only a few quick wing beats are needed for a fly to pitch and roll away from danger. A surprise in this examination was the high-speed rotations, which out-paced regular flight, a fruit fly used considering they are equipped with such small brains and musculatures.
The next step in the research was understanding the physics in their sharp turns. Fruit flies are tiny and taking in all the details on such a small creature proved frustrating. In order to study the mechanics of flight, the fly’s banks (turns), in detail, a giant robotic fly with a two-foot wingspan was constructed. Most researchers scale down models because of cost or other considerations, Dickinson, on the other hand, needed to see the wings up close and large.
Using mineral oil, a fairly dense liquid, and immersing the robotic wing in it, a slow flap of the wing produces the same effect generated by a tiny insect in mid-air. Scientists struggle to construct and study tiny, reliable robots in flight, and Dickinson’s research may offer more insightful information in micro-robotics. Knowing people are working to produce insect robots, Dickinson exclaimed engineers and scientists using his data may bridge a gap between the disciplines of biologists and engineers.
The next challenge for Dickinson is the study of wind gusts, odors and how the information is rendered into skillful flight. Another path concerning flight may not lie with insects at all, research conducted at the University of Minnesota is using the power of the mind to control flight, though for very different reasons. Professor Bin He, a biomedical engineer is using brain waves to control the flight path of a flying robot, causing it to dip, turn, fly higher, and also sail through a ring.
Using brain waves conjures up images of mad scientists controlling hordes of people. However, the type of science He is using is completely non-invasive. Electroencephalography (EEG) waves are sensed by electrodes in an EEG cap on the head. This type of technology, introduced by He, will possibly permit those with mobility or speech problems from neurodegenerative diseases recapture function of artificial limbs, wheel chairs or other devices.
Walking, drinking a glass of water, opening a door, all require thinking about the action and signaling the brain to make it happen, neurons in the motor cortex activate electric currents sending the proper signal and causing the required movement. He mapped the motor cortex, the area of the brain that controls movement. Sorting out this assortment of signals was the floor plan for the brain-computer interfaces (BCI).
Merely thinking about a motion, such as making a fist, offered the most easily recognized signals, compared to tying a shoe lace, which requires many more signals sent to the brain. Wearing the EEG cap and thinking about an activity sends a signal to a computer, which translates those signals into an electronic command. Volunteers were taken through stages of commands until they could use a three-dimensional object, a flying robot, just by concentrating on an action.
Dickinson, the biologist at the University of Washington, is also looking into brains, fruit fly brains, from a different angle. Next on his agenda is the study of the circuitry of fruit fly brains to discover the particular mechanisms behind their flight and apply this knowledge to future robots.
By Andy Towle