The Quadrotor’s Coming of Age

In the last few decades, fixed-wing Unmanned Aerial Vehicles (UAVs) have become more popular and more commonly used for a variety of applications. Militaries around the world use UAVs for routine surveillance and to carry out basic attack strategies. Civilian applications include tasks from border protection to search and rescue. However advanced fixed-wing UAV technology is becoming, there remains the need for craft with greater maneuverability and hovering ability.

Helicopters have been able to fill the niche of full-sized, manned vehicles. Now, a new type of Micro Air Vehicle (MAV) is emerging as the small-scale equivalent of the full-sized helicopter. This new craft (see Fig. 1) is commonly referred to as a quadrotor. While the quadrotor is also rotor-based, it has a few key differences that set it apart from traditional helicopters and make it particularly attractive as an MAV. A four-rotor design allows quadrotors to be relatively simple in design yet highly reliable and maneuverable. Cutting-edge research is continuing to increase the viability of quadrotors by making advances in multi-craft communication, environment exploration, and maneuverability. If all of these developing qualities can be combined together, quadrotors would be capable of advanced autonomous missions that are currently not possible with any other vehicle.

In general, a quadrotor is a type of rotorcraft that uses two pairs of counter-rotating, fix-pitched blades for lift. The use of fixed-pitched blades allows quadrotor propellers to often be connected directly to four individual motors without the need for complicated linkages that control pitch. These motors are then connected in an ‘X’ configuration. To power and control the rotors, a battery and microcontroller are placed near the center of the craft. Changes to the altitude and attitude (the height and orientation with respect to the ground) of the craft are achieved by varying the speed of individual rotors. Fig. 2 shows the typical layout of these components and identifies the different rotor changes needed to adjust the state of the craft. With such a straightforward design, it is easy to build vehicles that are much smaller than traditional rotorcrafts. Many of the quadrotors currently in development have a total system weight of about 1.5 kg and a rotor tip to rotor tip length of 0.5 m. Because of these small dimensions, quadrotors are better suited for indoor and outdoor urban environments than most other air vehicles.

Such a machine has a few additional benefits over traditional helicopters. With a traditional helicopter design, a large primary rotor is used to generate lift, and changes in thrust are generally achieved by varying the pitch of the rotor blades because changing the speed of one large rotor requires too much time. Furthermore, the mechanical parts needed to adjust the pitch of the fast-spinning blade are complex and difficult to maintain. Quadrotors, however, can achieve thrust changes by varying the speed of each of the smaller and lighter rotors, allowing the complicated variable pitch components to be avoided. In addition, because of the single primary rotor, a traditional helicopter must have a tail rotor to counteract the torque created by the primary rotor. Quadrotors, on the other hand, do not need a tail rotor, since the counter-rotating rotors cancel out each other’s torques. These differences reduce quadrotors to a vastly simpler design that is cheaper to build and easier to maintain.

Even with such a simple design, building a quadrotor poses some difficult challenges. In particular, quadrotors are hard to control and can tip over easily. Because its mass is concentrated in a small area, the rotors must react very quickly to counteract the tendency to tip over. An inertial measurement unit (IMU) is often used to detect this tipping. IMUs are made up of accelerometers and gyroscopes that detect the quadrotor’s behavior.

When the quadrotor begins to tip to one side, a microcontroller receives a signal from the IMU and calculates new motor commands that will compensate for the tipping motion. These motor commands are then sent to the motor controllers, which adjust the speed of their respective rotor to the proper degree. At this point, a corrective force returns the quadrotor to a level position. In practice, this whole process needs to happen at least 100 times per second to keep the quadrotor in the air. Motors and electronics that could update at this rate were not readily available until a few years ago [1]. Now that these products are available, small-scale quadrotor development is taking off.

While seen as a physically mature platform today, quadrotors went through numerous design phases during the 1900’s. In 1922, Etienne Oehmichen built the first quadrotor. The Oehmichen No. 2 was one of six full-scale rotary-winged vehicles developed by Oehmichen. This machine featured the familiar ‘X’ design that current quadrotors still use. However, since a single 120HP engine powered all four of Oehmichen’s rotors, it was not possible to control the attitude or orientation of the craft by varying the speed of individual rotors. To solve this problem, four additional small propellers were added to control the movement of the quadrotor [2]. In 1924, the Oehmichen No. 2 set the first Fédération Aéronautique Internationale distance record of a rotorcraft at 360 m [2]. Although this craft made great strides in the field of rotorcraft, Oehmichen was dissatisfied with the Oehmichen No. 2’s limited altitude. As a result, Oehmichen’s later iterations ditched the quadrotor configuration in favor of a single main rotor that could provide more thrust and allow higher altitudes to be reached.

It was not until 1956 that the next major quadrotor was developed (see Fig. 3). Convertawing’s Model ‘A’ quadrotor eliminated the extra four rotors used in the Oehmichen No. 2 and featured wings for additional lift in forward flight. The craft was powered by two engines and was controlled by varying the power to each engine [3]. Although this craft flew many successful flights while in development, the project was scrapped due to lack of interest from the commercial and military sectors. Even though the project was scrapped, Convertawing’s contributions live on, as this novel control scheme is now the primary control scheme for current quadrotors. From this point on, future quadrotor development and most rotorcraft research focused on improving the single primary rotor design until the technological advancements that allowed development of small-scale quadrotors came about.

The Stanford Testbed of Autonomous Rotorcraft for Multi Agent Control (STARMAC) project at Stanford University performed some of the initial work on making small-scale quadrotors autonomous. Stanford was able to modify commercially available quadrotors (DraganFlyer X4s) to follow a series of GPS waypoints [4]. After achieving this feat, Stanford created the STARMAC II series of quadrotors (see Fig. 1) with the goal of improving stability and control to make quadrotors super stable. To make their quadrotors more stable, the Stanford team focused their attention on modeling various aerodynamic effects that had previously been ignored. These included non-zero angles of attack, blade flapping, and airframe drag [4]. Non-zero angles of attack occur anytime a quadrotor moves and is not parallel to the ground. As the rotors spin, there is sometimes a slight bend, which leads to blade flapping. Airframe drag is the effect that the airframe against the flow of air. By taking these effects into consideration, Stanford was able to produce a quadrotor that was super stable, even in the face of disturbances such as wind. The development of such stable systems by Stanford’s team and other groups has allowed autonomous quadrotors to leave the research lab and enter the real world.

The largest use of quadrotors has been in the field of aerial imagery. Traditionally, stationary aerial imagery was only possible with full-sized helicopters. Quadrotor UAVs are a perfect replacement for this job because of their autonomous nature and huge cost savings. Capturing aerial imagery with a quadrotor is as simple as programming GPS coordinates and hitting the go button. The company called DraganFlyer currently sells video-capable quadrotors and various camera payloads for their X4 quadrotor. A picture of one of the configurations is shown in Fig. 4. The available payloads include a standard Panasonic digital still camera that is capable of 720p video, a video-only camera with Digital Video Recorder (DVR), a black-and-white low light video camera with DVR, and thermal infrared video camera with DVR. All of these video cameras have the option of being streamed live to the ground. With such diverse camera packages available, the X4 can be used in a number of different scenarios. These applications include everything from real estate photography to industrial systems inspection to military tactical surveillance. One company, Microdrones GmbH, is already seeing an increased use in its quadrotor as a surveillance tool. The German-made Microdrones are currently used in Germany, Belgium, and Norway, and pilot programs in England have been using these quadrotors since 2007 [5]. The Liverpool police use the Microdrone md4-200 quadrotor as an evidence-gathering tool and a fire brigade in West Midlands has also purchased an md4-200 for assistance during emergency situations. Both of these organizations are taking advantage of the quadrotor’s closed-circuit television capabilities and ability to provide an eye in the sky to the action below. The deputy chief fire officer of the West Midlands fire brigade says:

“This is fantastic new technology that will provide real benefits when we are tackling a range of emergency situations. Being able to look down on the scene will allow us to get a full picture of the incident and the surrounding environment, which will aid incident commanders to make vital, potentially life-saving decisions.PA News

With quadrotors continuing to gain acceptance as aerial imaging tools, the way has been paved for the introduction of new quadrotor applications.

Many new quadrotor applications will result from improvements in quadrotor collaboration. The MIT SWARM Health Management program introduces multi-quadrotor surveillance missions. One such mission has multiple quadrotors tracking and following moving ground targets. The health management feature of the missions allows quadrotors with low battery power or a critical failure to be automatically replaced with new quadrotors. In the case that a quadrotor’s battery is low, the quadrotor can land in a recharging station and then substitute for another quadrotor (“UAV SWARM Health Management Project”). Fig. 5 shows a snapshot of a SWARM demonstration. Such an effective health management system allows for continuous mission operation. The major limitation of this system stems from the need for a sophisticated high-speed camera system to communicate the positions of the currently operational quadrotors to each other. Newer projects can better estimate a quadrotor’s location and orientation with enhanced sensors and localization algorithms.

Another progressive technology has recently been developed at the University of Pennsylvania’s General Robotics, Automation, Sensing, and Perception (GRASP) lab. This lab has achieved “aggressive” maneuvers with their quadrotors, including flips, high speed-low clearance stunts, and perches on vertical walls and inclined surfaces [6]. Fig. 6 shows a quadrotor zipping through a narrow window. In order to do this, the quadrotor must deviate from a stable orientation while traveling through the window. While this feat is impressive, it is currently only possible for preconceived motions and is not yet capable of calculating moves for new environments in real time. Also, like the MIT SWARM lab, the GRASP lab uses high frequency cameras to provide the quadrotor with information about its position. Such a system provides very accurate stabilization but cannot occur outside of a lab. Still, this novel research should pave the way for more robust, advanced maneuvers.

With all of these technologies on the horizon, a variety of new applications will appear. Aerial imaging capabilities will increase with the ability for multiple quadrotors to interact together in the same space. Quadrotors could be used to scan out a large area in synchronization, or image one area for a longer period of time with a health management policy. New venues, like sporting arenas, might also be able to take advantage of more robust aerial imaging. As the technology matures, Americans might also begin to see quadrotors in their hometowns, as is already possible in Europe. The police and military can implement these technologies in advanced tracking and surveillance scenarios. Imagine a quadrotor being used to chase a suspected criminal through dense city streets and into buildings, all without any prior knowledge of the environment. No existing machines can accomplish such a task. As research continues to push the boundaries of quadrotor capabilities, these and many more applications will unfold.

Quadrotors have a come a long way from the original Oehmichen No. 2 design of the early 1900’s. Improvements to small-scale sensors, motors, and microcontrollers have revived quadrotor development by bringing older designs to a smaller scale. While already in use as an aerial imaging tool for private, commercial, and government use, new research is expanding the possibilities of quadrotor applications. Advances in multi-craft communication, environment exploration, and maneuverability will lead to advanced autonomous missions that are currently not possible with other any other vehicle.