A simulator helps develop and test autonomous behaviors prior to flight. Photo: Aurora Flight Sciences
AI’s Impressive Impact
To prove his point about AI’s impact, Cavanna described the many ways that AI is now being used at Leonardo Helicopters. In engineering, for instance, Leonardo uses AI to test and verify the validity of its designs. The company also uses AI to improve its manufacturing processes, management systems, and training and simulation offerings.
“AI can help speed up the training process by recognizing the maneuvers that students are making in the simulators and giving them scores quickly,” Cavanna said. “It is also useful in preventive maintenance where AI can analyze the data being compiled by ongoing, sensor- based health monitoring of the aircraft and use it to project what maintenance needs to be done. Meanwhile, we are using generative AI to make sense of vast amounts of documents. AI is also important for autonomous flight, such as the autonomous uncrewed projects that Leonardo is doing with the U.K. Ministry of Defence and the U.S. Marines.”
Sikorsky, a Lockheed Martin company, is constantly evaluating the use of AI and “machine
learning” on its helicopters
and future VTOL products, both in the lab and in the air. What’s the reason for this two-pronged approach? “Testing and
qualifying AI (machine learning)
for applications to direct aircraft control is difficult,” replied Igor Cherepinsky, director of Sikorsky Innovations. “That’s because of the inability to test and verify these algorithms accurately in the lab alone.”
76 Sept/Oct 2024
Nevertheless, “Sikorsky can and does use AI in adjacent roles,” Cherepinsky noted. “For example, our Matrix flight control system is capable of performing full-mission planning from a handful of goals and constraints. Once the system comes up with several plans, we use reinforcement
learning algorithms to
pick a plan that is more desirable for the human crew. In this case, the AI algorithms are not performing a flight- critical function. So, let’s say I’m in a Black Hawk helicopter performing a military cargo supply mission. AI would pick which routes one or more aircraft need to make on the battlefield to get from Point A to Point B, and with which helicopters. Another time, the algorithm might choose a different solution to get supplies to the troops. Either outcome is safe.”
Over at Aurora Flight Sciences, AI is having a big impact on how the company designs and builds special-purpose drones. “In aircraft development, AI is advancing how engineers solve problems and deploy technical capability,” said Church. “Simulations, modeling, and digital twins have evolved to give a more complete analytical simulation. AI means we can feed in more constraints and simulate every option of every variable, generating a full cloud set of data points.”
AI is also making Aurora Flight Sciences’ drones more successful in fulfilling their missions. The reason: “In our flight operations, the strategic integration of AI-enabled features allows us to improve safety, reduce pilot workload, and increase data quality during flight
Of course, the use of AI does vary considerably from company to company.
JumpAero.com says Jump Aero is “bringing to market the fastest electric- vertical-takeoff-and-landing (eVTOL) aircraft for first responders.” According to company founder and CEO Carl Dietrich, “We do not use AI for anything that could be considered safety critical, which is most of what we do. However, we do have experience working under a USAF research contract with Caltech on an adaptive/ML flight control routine that demonstrated meaningful improvements in the response of the flight control system to failures. Importantly, however, the system is trained in a simulated environment. As such, the AI/ML component is static after training. This is important for certification and general confidence/reliability.”
Having said this, Dietrich acknowledged that “Other companies that are in larger volume production use AI/ML as part of continued airworthiness and quality control/inspection process. Vision-based algorithms are useful in this context, as well as in advanced autonomy routines.”
testing,” Church said. “AI can allow the flight crew more capacity to focus on the system being tested. We also see quality improvements in that we can ensure that tests are repeatable, which means we are analyzing the technology rather than the variability in human operators.”
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76 |
Page 77 |
Page 78 |
Page 79 |
Page 80 |
Page 81 |
Page 82 |
Page 83 |
Page 84
