Compare 2 Drone Pilots: Flying High with Analytics


In recent years, an increasing number of government and law enforcement agencies use drones to get a bird’s eye view from the sky. To be effective, drones must be able to maneuver quickly to their intended target. Swift maneuvers enable them to navigate complex environments and accomplish their tasks efficiently. The requirements to become a certified drone pilot vary by country, but all require that you pass a written exam on aeronautical knowledge administered by the aviation authority. Surprisingly, (at the time of this post) there is no requirement to demonstrate performance. So how can we compare the flight skills of two certified drone pilots?

Scenario:

Two certified drone pilots, Aleks and Dustin, are evaluated for how quickly they position the drone at a specified target, in this case five hundred feet away from the launch position.

We want to compare to see if Aleks and Dustin’s performance is different from one another. The time is measured in seconds when they start (includes lift, movement to target). A practical difference is 0.5 seconds – meaning that if one pilot can get the drone into position ½ second faster, they are better. Here is the experimental data:

Picture1 (4)

Using Minitab Statistical Software, we can quickly create descriptive statistics.

Picture2-Nov-30-2023-07-42-24-3932-PM

We see that across ten flights, Aleks’ average time to target is 15.86 seconds, while Dustin’s is 12.42 seconds.

drone estimate

The estimated difference is 3.4 seconds. The 95% confidence interval tells us that the true average difference between the pilots is as small as 1.736 and as large as 5.131 seconds. Based on this confidence interval, we can confidently say the flight times are not only different than zero, but because the lower end of the confidence interval for the difference is more than our practical difference of 0.5 seconds, we conclude Dustin is the better drone pilot in this test.

A boxplot of the flying time provides a visual of the analysis:

boxplot

 

In the future we could perform follow-up experiments to include multiple distances, obstacles, varied flying conditions and multiple drones for the pilots to use. Minitab Statistical Software provides analytics and visualizations, making it easy to measure performance.

Experiment with your own data analysis – get a free trial of Minitab Statistical SoftwareStart 30-Day Free Trial





Source link

Be the first to comment

Leave a Reply

Your email address will not be published.


*