Artificial Intelligence and Training

Artificial Intelligence and Training

How can algorithms, machine learning, and artificial intelligence make a more meaningful Unmanned Aerial System (UAS) training experience? Data and analytics of UAS pilots’ individual performance, whole populations, or subsets of populations- like specific industry, can contribute to a more streamlined training experience. A training provider can preemptively offer courses, should a deficiency exist, or proficiency advance learners based on experiences or that of similar UAS pilot profiles. The advancements of algorithms, machine learning, and artificial intelligence are genuine contributors to the future of UAS training.

The analysis of data gathered on students for the purposes of creating more efficient and more effective training allows a training provider to identify problems and address them much earlier in the training process. This can provide the necessary time to undo negative learning or fix poor technique and allow ample time to apply new learning under the direction of an instructor. It can be a much more positive experience for the student with better learning outcomes. Results from student-specific data analytics are added to a student’s profile. This would allow the student to begin a training event that is adapted to his/her unique challenges and capabilities. The student’s data and resulting analytics are then continually tracked going forward, developing a more robust profile of the student’s learning style, trends, and proficiencies.

When an instructor is monitoring more than one student, and when a student decides to take a shortcut, or forego procedures, or physically does not have visibility of the causal factor for the poor performance, it may or may not be caught by an instructor. 

With data collection and real-time reporting, anything outside of an instructor’s view or evaluation ability can be captured via machine (computer). The instructor can review and verify the information and correct the behavior before bad habits are instilled. As an example, if a student skips a step on a preflight inspection, like spinning propeller blades for unobstructed movement, and then has a flight scenario (real or simulated) where the propeller blades malfunction, the teachable moment is evident; skipping steps, or taking shortcuts can yield disaster. Eye movements and scan patterns are a highly subjective point of instruction. A seasoned instructor understands a new learner has a poor scan pattern. AI would allow detailed evaluation and data to not only offer the learner with productive feedback but assist the instructor in opportunities to help the learning process. The instructor would have undisputable feedback for the learner to improve the outcome and reach the source of the errors immediately. We have observable cases where the causal driver of poor performance was a variable that a human would not be able to capture, but a machine is able to very rapidly root cause, review all available variables and determine those that were statistically significant drivers of the poor performance, and offer objective data to establish improvement areas.

"When Experiencing Flight Training, It Can Be Rewarding And Valuable To Be In A Class With Pilots Of Many Different Backgrounds, Experience Levels, And Industries"

When experiencing flight training, it can be rewarding and valuable to be in a class with pilots of many different backgrounds, experience levels, and industries. Facilitated discussions can increase the value in learning at any level, including an instructor’s. A learner’s unique practical skill level and ability should guide a more challenging and rewarding lesson with tailored learning objectives based on specific desired outcomes of the company or agency which employ the learner. Instructors that can access real-time, comprehensive objective and subjective data, like aviation experience, UAS experiences, aircraft types, industry work, previous training events, and more, can tailor training scenarios to best challenge and evaluate the student to the highest standards. These captured data points and analytics will continue to feed the machine and provide objective evidence and evaluations of the student's performance, thereby differentiating the training experiences for future training events.

Why are we interested in exploring this trend in algorithms, machine learning, and artificial intelligence? Every bit of data is important to training. If someone lacks sleep, hydration, nutrition, or experience, the training received is impacted. Data and analytics can increase training effectiveness and enhance safety through automated, intelligent, and objective training.  

Weekly Brief

Read Also

Optimising Predictive Maintenance for Airlines

Enea Fracassi, Chief Operating Officer,Corsair International

Predictive Maintenance

Jeff Moore, Sr. Director, Line Maintenance and Technical Services, Frontier Airlines

The Fast Moving Space Industry is Embracing New Methods

Elizabeth Seward, Head of Strategy – Space, BAE Systems

The State Of Automation In The Aerospace Industry

Pedro Imaz, Head of Manufacturing, Héroux-Devtek

Future Of Space Security: What The Industry Can Do To Improve Cyber Resiliency

John Fant, Vice President of Advanced Program Development, Lockheed Martin Space

The Surveillance Challenge

Sven Rensmeyer, Military Surveillance Expert, Cmc / Cns Unit, Eurocontrol