In order to create value with a drone program, being able to successfully fly the thing is obviously crucial. Whether you call it an autopilot, a co-pilot, a ground control station, command and control link, or any other term, this key mechanism is how to tell the aircraft where to go and what to do. Like many problems in the new and exciting world of drones, there are dozens of ways that you can handle this. We have broken the broad group of control methods down to three categories, summarized below. The important question is figuring out what works best for your application.

Full UAV Autopilot

Description: Aircraft is controlled by a computer from takeoff to landing; no human intervention required
Use Cases: Large, flat, open areas
Advantages: Very simple, low chance of human error
Disadvantages: Limited applications, limited collision avoidance, limited camera control

Computer Assisted Flight

Description: Pilot controls the aircraft, with significant assistance from a computer / GPS guidance system
Use Cases: Houses, buildings, complex structures
Advantages: High situational awareness, prevention of many common errors
Disadvantages: Some training required, higher risk of human error / collision

Manual Flight

Description: Little or no computer assistance; pilot has full and complete control of the aircraft
Use Cases: Performance flying, racing
Advantages: Can perform acrobatic maneuvers; can push hardware limitations
Disadvantages: Extremely difficult to master; crashes & errors common, even among well-trained operators

Full UAV autopilot means reduced training needs

Full autopilot involves the computer takes control of the aircraft for the entire operation, including takeoff and landing. These systems navigate using a mix of technologies, usually relying heavily on GPS. However, very few commercially available autopilots are able to reliably fly anything beyond a simple grid pattern. Further, few of these technologies are able to actively see and avoid obstacles. Because of these limitations, full autopilot is really only useful in large, open areas, where a grid pattern survey of the area is useful. In these cases, however, full autopilot is highly effective. By minimizing the need for pilot intervention, the training requirements for operating UAVs drops significantly, making starting a drone operation cheaper and easier than with alternatives.

An example of an automated flight plan pattern automatically created by DroneDeploy, a provider of autopilot software.

Computer Assisted UAV Flight means pilots focus on mission goals

Where full autopilot is not enough, most current UAVs are capable of some type of computer assisted manual flight. In these cases, the pilot still controls the aircraft via remote controller navigation sticks. However, the autopilot system adds some significant assistive features. The most valuable of these is "position hold," which locks the aircraft in position when the pilot releases the controls. This key safety feature prevents many of the most common mistakes that can lead to accidents. Other computer assistance features include auto level, which is included in nearly all drones these days, an automatic return-home feature, and orientation control as well, which can all be useful in certain circumstances. Ultimately, all of these features are meant to make the pilot's job easier, allowing them to focus more on the data they are trying to collect. Computer assisted flight ideal for taking photos of houses, bridges, or other structures for which a top-down viewpoint is not enough.

Full manual UAV flight is for experts only

When you take away all of the computer assisted features of most current commercial drones, you are left with full manual flight. Those that used to fly conventional RC aircraft know the challenge associated with full manual flight, in which a slight error can quickly send your aircraft plummeting into the ground. The level of training and experience that it takes to become proficient in full manual flight is incredible, with hundreds or thousands of flight hours needed to become highly skilled. However, once proficient, full manual flight is truly a sight to behold. It allows the operator to perform flips, curves, inversions, and all kinds of maneuvers that any sensible autopilot would not allow. The only practical use case for manual flight is drone racing, which is certainly impressive.

Which type of UAV piloting is right for you?

If you are thinking about using drones in your business, full autopilot technology is likely the best way to go, so long as you understand the limitations. Even in the case of full autopilot technology, we always recommend basic pilot training so that in case of an emergency or hardware failure, you can still safely bring the aircraft back home. Assisted manual flight presents the same scenario. While highly useful, it is still possible to crash without the proper training and experience. And full free flight is only for those who want to pursue racing. Although each style of autopilot has its positives and negatives, ultimately the right autopilot is the one that matches the needs of your operation.

Have any questions about which pilot style is right for you, or how to best integrate autopilot technology into your business? Leave a note in the comments below or drop us a line at info@aerotas.com.

Logan is the founder and CEO of Aerotas. He graduated from Harvard Business School in 2015 and brings experience in a broad range of industries, including agriculture, real estate, and finance. A lifelong tinkerer and avid UAV enthusiast, Logan also has hundreds of hours of flight experience over a broad set of different UAV types.