The systems comprising an MAV include the airfoil or lift system, a propulsion or energy source that meets airborne and ground movement needs along with the power requirements for support systems, control surfaces to meet a flight profile, sensors specific to mission requirements, a microprocessor that allows the MAV to fly autonomously and collect information, and a communications system that allows the MAV to transmit or store (for subsequent downloading) collected data.
MAV airfoils are relatively small and subsequently sensitive to anything other than perfect wind conditions. Wind gusts, rain, snow, and even steady high wind conditions can cause an MAV to fail catastrophically or at least not meet its mission requirements.
At these small sizes, the loss of lift due to the detachment of an aerodynamic boundary layer on a airfoil brought on by a maneuver-induced stall is exceedingly hard to recover from by conventional means. Redesigned airfoils/control surfaces are required to adjust for these conditions.
Propulsion systems like internal combustion or jet engines that might be used to drive propellers for MAVs become difficult to cost-effectively produce due to the manufacturing tolerances and weight considerations required. The weight of the propulsion system and its store of fuel also become an increasingly larger portion of the overall vehicle weight.
Nevertheless, researchers are studying such concepts as tiny jet engines, ducted fans, and pulse jets. Pulse jets are a favorite since they have no moving parts and could provide air for lift and flight control.
Off-board energy sources are also being considered. In one system, microwave energy from a surface search radar system could be beamed to the MAV. The microwave energy would then be converted to a dc signal to drive an electric motor for driving a conventional propeller.
Sensory systems are one of the few off-the-shelf technologies that designers will be able to use with few modifications. A selection of sensors that might be used on MAVs include imaging devices, chemical monitors, inertial measurement systems (for autonomous flight operation), range finders, force/torque sensors, sonar, position determination sensors, linear motion measurement systems (for ground motion), interfacing sensors, and accelerometers. Additional sensor-like devices that may be needed on an MAV include pan/tilt mechanisms and shape-memory devices.
Many of these devices either already exist or are being developed in thin-film or microelectromechanical systems (MEMS) configurations under separate research programs.
MAV microprocessors may also be chosen from a selection of existing devices. Overall processing capabilities are not an issue, the amount of information needing to be processed is less than current devices are able to handle. Microprocessor size and associated hardware, however, may be an issue, and devices that utilize 3-D stacking technologies could be utilized. Stacking a microprocessor, memory system, and sensor into a package smaller than a sugar cube is already a viable technology.
Software used with a microprocessor for managing sensor information is relatively straightforward and exists for larger surveillance vehicles.
The primary use of the microprocessor, however, would be for flight control. Flight control systems for conventional aircraft utilize linear, time invariant control theories. Adaptive wings that might be used on MAVs, combined with unstable wind conditions, make control solutions nonlinear and time-variant.
Software for managing the control surfaces, mission flight profile, and avoiding unforeseen obstacles requires a substantial degree of built-in intelligence and “robustness.” Artificial intelligence techniques such as fuzzy logic, neural networks, knowledge-based systems, genetic algorithms, or combinations of all may need to be integrated into a final configuration.
Control systems will determine whether the MAV is a practical concept. Most researchers agree that some type of adaptive control system will be needed for a flyable UAV. Several types are being evaluated.
One system being developed by Ron Barrett at Auburn Univ., Auburn, Ala., uses piezoelectric torque plates to actively twist or deform an airfoil in response to an applied electric field. Another control surface concept employs piezoelectric bending elements to pitch an airfoil shell about the main spar.
Still another system being developed by Robert Englar, a principal research engineer at Georgia Tech Research Institute (GTRI), Atlanta, combines an altered wing design with engine exhaust blown out through channels in the wing’s trailing edge. This concept, known as the Coanda effect, results in increased lift and control capabilities up to four times those of conventional control surfaces.
With all of these systems to consider, researchers are tackling the most fundamental problem first, that of developing integrated powered flight systems.
Robert Michelson, a principal research engineer at GTRI, has taken an unconventional approach to development of a MAV, by creating a robotic flying insect, or entomopter, as he refers to it.
Michelson’s entomopter is designed to be used indoors. “There are far too many problems that may never be solved in developing a cost-effective MAV for low-speed flight in unstable wind environments,” he says. While the air currents in a building are more predictable, the enclosed area makes object avoidance a larger issue than it might be outdoors.
The applications for interior surveillance are equally useful for appraising military operations, finding drug labs, monitoring hostage situations, and evaluating hazardous environments like leaking nuclear reactors.
Michelson’s entomopter departs from the conventional designs by incorporating flapping wings as its propulsive mechanism. “We can obtain lift on both the up and down strokes of a flapping wing,” he says.
“The smaller the vehicle, the less reasonable is a fixed-wing design because fixed-wing vehicles rely strictly on lift generated by airflow over a vehicle moving through the air to support the weight of the vehicle,” says Michelson. “The smaller the vehicle, the less lift it can supply. Most vehicles counter this effect by increasing the velocity of the vehicle, which is unacceptable in situations such as the indoor missions where MAVs would make the most sense.”
Michelson’s propulsion system also departs from the conventional, noting that MAV operation necessitates use of the highest possible energy source at the lowest possible weight. Gasoline is a high energy source, having a much higher energy density than similar weight and volume high density batteries. Michelson’s propulsion system would process gasoline in a chemical reactor at a rate tuned to the power needs of the flapping wing propulsor.
A typical flight mission envisioned by Michelson is to fly his entomopter a short distance within a building and land somewhere near the final destination. The robot would then crawl on the ground to the final destination and drop its monitor or sit with its monitor in place in a powered-down mode for an indefinite period of time or until some object or person passes by to activate it.
Michelson has demonstrated proof-of-concept devices for his entomopter and expects to fly a full-up mockup in straight level flight this summer. If he obtains DARPA funding this year, he could have an entomopter in fully controllable flight missions within a year.
Ephrahim Garcia and Michael Goldfarb, professors of mechanical engineering at Vanderbilt Univ., Nashville, Tenn., recently received a three-year DARPA contract to produce robotic insect crawlers for transporting video monitors into remote or hostile environments. They also are looking into an insect-based flapping wing design – only theirs is powered by piezoelectric actuators.
A few commercial companies are also involved in MAVs. Researchers at AeroVironment, Simi Valley, Calif., are developing a 15-mm dia flying saucer design with a battery-driven electric motor powering a conventional propeller and three smaller motors powering flap actuators.
MLB Co., Palo Alto, Calif., has also developed a 10-mm MAV prototype Flyswatter that sits on its tail to provide vertical take off and land capabilities.
Michelson feels confident about the capabilities of his entomopter design. “A lot of research has been done on insect flight and the mechanisms are now fairly well understood,” he says. “We also have enough energy in a small enough size to make it work.”
Making Robots Fly
Robert Michelson’s reciprocating chemical muscle (RCM) is at the heart of making his robotic flying insect work. The RCM is a regenerative device that converts the chemical energy found in conventional fuels, such as gasoline, into motion through a direct noncombustive chemical reaction.
No combustion takes place nor is there any ignition system. The reaction is anaerobic allowing it to operate underwater or in oxygen-starved environments and gaseous by-products of the reaction can be used for gas-blown control surfaces.
A byproduct of the reaction is electrical energy which can be used in onboard control and microprocessor systems. It is generated thermoelectrically from the reaction’s own exothermic metabolism.
Michelson, a principal research engineer at Georgia Tech Research Institute, Atlanta, feels that his RCM system is also directly scalable into microelectromechanical systems (MEMS), providing even further reductions in weight and size in future designs.