There are many practical applications for micro UAV’s. In cases such as military reconnaissance or emergency response, instant availability of high quality imagery can have life or death consequences. In other cases such as aerial surveys or environmental monitoring, workers have limited time and resources to successfully complete a job. In all cases, the absolute necessity of the UAV operator is to quickly and efficiently provide the high quality data that is needed in the field.
Unfortunately with many micro UAV’s, collecting the data often takes a back seat to stabilizing and controlling the vehicle and its payload. Rather than specifically focusing on the priority task of obtaining quality imagery, the UAV operater spends more time focused on controlling the system and its payload. This whitepaper outlines why an intuitive control methodology is key to the success of a UAV program and what key attributes to look for in a system.
There are many critical uses for Small and Micro UAVs in tactical military, law enforcement and civilian operations. For tactical military missions, small groups of soldiers rarely have instant access to high-level UAV intelligence and often approach an area by foot or a single vehicle. This means to quickly obtain aerial intelligence, they must carry a small UAV in a portable kit, deploy and operate the system themselves and collect intelligence locally.
Law enforcement applications can include emergency response, search and rescue, forensics, traffic accident reconstruction and crowd control. Civilian applications can include GIS mapping, environmental monitoring or inspection services. Typically, these applications are localized, have limited budgets and the operators are expertson collecting data, not expert UAV pilots. Users typically arrive on scene in a car, and require quick deployment with the resources available at the scene.
Regardless of the application, the goal of deploying a Micro UAV is to quickly and efficiently provide critical aerial based information to precisely where it’s needed – to the people in the field. In order to make micro UAVs a practical solution for these applications, the following must criteria must be met:
- Skilled operators must be readily available at all times to obtain aerial imagery in a timely fashion
- Operator training requirements must be minimal to quickly increase the pool of skilled operators
- Piloting the micro UAV cannot be complex, allowing non-pilots and typical field personnel the ability to quickly become qualified pilots
- Tactical micro UAVs operations must require only a single operator for both flight operations and controlling the payload, allowing the majority of field personnel to focus on the task at hand
- The micro UAV must contain automated safety features and environmental controls (such as wind and rain) allowing typical pilots to focus on aerial imagery
While virtual cockpit style controls are familiar for users of large unmanned aerial systems, smaller systems often find their roots in the hobby world. As a result, most small UAVs are controlled using R/C-style joysticks, augmented to control additional aspects of the UAV, such as its payload or display a video downlink.
Control of the vehicle, therefore, requires a dedicated operator, with a second dedicated operator required to obtain the imagery. This operator requires a great degree of hand-eye coordination and significant training in order to successfully pilot the system. Manufacturers of R/C-based systems realized the difficulty of piloting with joysticks and have added stabilization electronics. This stabilization help keeps the craft level and is designed to prevent it from rolling over, but doesn’t control the vehicle’s position or keep it from flying away.
One Operator, Too Many Controls!
A typical joystick-controlled system has 2 joysticks. One controls the motor power and the rotation (yaw). The other controls the vehicle pitch and roll. To hover in place, an operator must find the proper balance of motor power, roll, pitch and yaw controls. Too much power and the vehicle climbs, too little and it descends. Too much yaw and it spins in a circle. As one might imagine, it takes a significant amount of time and dexterity to learn and control these 4 individual variables at once.
When the vehicle is facing away from the operator, pushing left on the stick, makes the vehicle move left. However, when the vehicle is facing towards the operator, all the controls become reversed, requiring the operator to mirror all the inputs. With all this thinking and training required, the operator is no longer focused on taking photographs or collecting data – all their energy is going into just keeping the vehicle in the air.
Adding wind, rain or difficult viewing conditions and it becomes extremely difficult for an operator to successfully pilot the craft. While lights on the craft may increase the visibility at night, the difficulty of piloting the craft increases exponentially in the dark with only a few points of reference.
It is estimated that learning to fly an R/C helicopter requires approximately: [Source: electric-rc-helicopter.com]
- 30 hours of practice to achieve a stable hover; and
- An additional 30 hours to achieve stable forward flight.
This only yields a pilot with basic skill level. In order to pilot for a mission at a distance, in any environment and near obstacles, an operator must learn to compensate for wind and vehicle orientation, build additional skill to fly at a distance and in difficult visibility and then control a payload. This skill does not come easily to most people and for some is not practical to learn. This makes it difficult to build up a large crew of capable pilots, which, therefore, inhibits large scale deployments of this technology.
In fact, “Over 50% of all single rotor collective pitch RC helicopters sold to and built by first time fliers end up crashing in the first few minutes of their maiden flight.” [Source: rchelicopterfun]
Line of Sight Required
Many operations require a UAV to be positioned beyond line-of-sight, at high altitudes or operated in the dark. For example, in a military reconnaissance scenario, soldiers may deploy the craft a significant distance from the target and be required to operate at high altitudes to avoid detection. With a joystick operated system, the operator must have a clear view of the UAV at all times. In difficult conditions it can be easy to lose sight of the craft, or become confused about its orientation.
Even in ideal environmental conditions this complexity of control with joysticks requires a secondary payload operator to steer and operate the payload. Not only does this payload operator have to control the camera, they must also have clear communications with the vehicle pilot to ensure the payload is positioned appropriately. When in a stressful or covert situation, losing focus is problematic and communication can easily break down, compromising the operation.
An additional challenge of piloting an R/C controlled vehicle is the problem of perspective and depth perception. As a subject moves further from an observer, the ability of the observer to judge its distance decreases exponentially, especially when that subject is flying and there are no comparable ground reference points.
Take the situation depicted below, for example. Due to human physiology, the operator will have a far easier time judging the distance of the blue UAV than the red one, since it’s easier to perceive differences in distances at closer range due to larger visual disparity. Additionally, due to depth perception difficulties and the limited relative visual cues provided by objects in the air, he will have an easier time determining if the blue UAV is above the tree than if the red UAV is above the red tree. Knowing the absolute position of the UAV is important for a pilot to execute the mission and safely operate the UAV.
All of these difficulties make joystick-controlled systems unsuitable for many military, law enforcement and civilian applications.
A more functional alternative to joystick-based control is the use of a map-based controller. While an R/C-style system is piloted with joysticks, a map-based system uses an aerial map and allows the operator to control the vehicle’s position, instead of its orientation. This is much more intuitive to the operator, as the vehicle is responsible for determining the correct motor power, roll, pitch and yaw, the user just tells it where to go. Because the user knows exactly where the system is, there are no issues of perspective, line-of-sight, or depth perspective.
How It Works
In a map-based configuration, a pilot simply presses a take-off button to launch the vehicle, drags a slider to climb to altitude, and touches the desired position on the map to fly the UAV to that point. Instead of trying to control many flight variables, which is not only difficult and distracting from the operation, the user targets the vehicle with a simple, intuitive touch.
There are two styles of operation developed in Aeryon Labs’ patent-pending control: waypoint and ad hoc. In waypoint mode, the operator can plan a series of GPS waypoints to guide the vehicle. These waypoints are connected by a path, and the vehicle can automatically fly from one point to the next without any user intervention. In ad hoc mode, the user controls the desired position of the vehicle in real-time by pointing to the desired location on the map. The new position can be targeted simply by pointing at a different location.
Targeting the camera or other payload is just as simple. Simply touch the map to point the camera and the UAV ensures the camera orientation has the best possible view of the target. Once a target is acquired, without any user intervention, the UAV can keep the camera pointed at those coordinates regardless of the movement of the UAV.
The map-based positioning and camera pointing relieves the operator of all real-time control duties, allowing them to stay aware of their surroundings and the situation, focus on the imagery, and complete the task at hand.
New User Training
This style of control has numerous advantages over traditional R/C techniques and has a substantially shorter training time, resulting from the significantly lower operator skill level required. In new-user trials, operators with no previous R/C or UAV experience were able to pilot the Aeryon Scout UAV with mere minutes of training. Since the vehicle itself does the control, a new user can fly in the very same weather conditions as an experienced pilot, with no need to learn to compensate for wind or other conditions. This allows the user to focus on the operation, not flying the vehicle.
Since the user is controlling the position of the vehicle with a map, and not its orientation, there is no need to see the vehicle. This means the vehicle can easily be flown beyond line-of-sight, at high altitudes or at night. The problems of depth perception and perspective are eliminated since the user can see the location of the vehicle on the map at all times. In the above example, in order position the UAV above the tree, the user would simply click above the tree in the map.
While a map-based interface provides solutions to the shortcomings of a joystick system, it also yields some additional benefits. Because of the additional on-board intelligence required on-board a map-based UAV, it’s able to respond intelligently to high winds, low battery and loss of communications situations. With a joystick-driven craft, the pilot must try to estimate wind speeds from the ground, making it difficult to know when it’s too windy at altitude. With no telemetry data from a standard R/C system, there is no way of knowing how much battery or fuel is left on the UAV, creating a dangerous situation. In a loss of communication situation, a standard R/C system will often fall from the sky, or simply fly away, while a map-based system can fly home and land automatically.
A very significant benefit to a map-based control architecture is the ability of a single pilot to control multiple vehicles from a single controller. With an R/C system, that is simply impossible. The use of multiple vehicles by a single pilot affords advantages to the operation such as reducing man-power requirements, permitting constant surveillance and covering a large search area quickly.
For example, in the scenario shown below, a single operator can send the blue UAV in to replace the red UAV once its battery is low and recall the red UAV. Each UAV manages its own safety and handles weather conditions independently and simultaneously.
While power sources are continuously improving, every micro UAV has a usable life span which may require it to return home before a surveillance is complete. This ability to dynamically replace UAVs in flight, allows the operator to keep one UAV on target at all times, providing persistent surveillance. All of this, with only one operator, and a very light training load.
If the goal of operating a UAV is mission-focused, such as survey, police work or military reconnaissance, a map-based system provides the functionality required to quickly train operators, deploy and deliver results. Joystick-driven systems have high training overhead and have significant operating challenges including multiple operator requirements, user dexterity, and line-of-sight limitations. A map-driven system is much more intuitive to the user, quickly learned and is a key to the success of a UAV program.