U.K. and Canada Push Startups to Address Aerospace Ice Protection with AI

Large corporations outsourcing research and development (R&D) to acquirable startups has always been a thing. Now it’s happening to artificial intelligence (AI) technology.

A new “ AI Innovation Challenge  ” tasks United Kingdom (U.K.) and Canadian startups and researchers to pitch ideas for how AI could improve aircraft efficiency in extreme weather. The challenge – organized by the UK Science and Innovation Network (SIN)  in Canada, London-based Digital Catapult  , Montreal-based Bombardier  , and the Consortium in Aerospace and Research & Innovation in Canada  (CARIC) – will encourage skills sharing and trade opportunities between the two countries while attracting highly talented and motivated individuals.

The challenge – funded with £30,000 ($38,754) from the UK Department for Business, Energy, and Industrial Strategy  and £70,000 ($90,426) from sponsors – is open to startups, scaleups, research institutions, and academic bodies with expertise in fields such as computer vision, generative models, predictive modeling, and physical simulations. The teams will showcase ideas as to how AI could make aircraft more aerodynamic and cut down on ice build-up.

Ice build-up occurs where water droplets colect on an aircraft as it flies through clouds. This surface water can take the form either of a microscopic film coating, or of small, fast-moving beads and streaks called rivulets. If left unchecked, hard, translucent ice shapes called “glaze ice” form on aerodynamic flight surfaces. Additionally, in colder conditions, instantly freezing droplets can collect on leading edges, forming white, opaque “rime ice.” These forms of surface water and ice may appear at the same time over different areas of an airplane on the ground and in flight.

Initial natural icing tests are used to determine how ice will accumulate on an aircraft during an actual flight, like on the Bombardier Global 7000  FTV2.

To improve aircraft efficiency, researchers will investigate how AI can be leveraged to better characterize ice buildup on aircraft. The development of ice protection systems involves detailed physical modeling and extensive testing in experimental facilities or in flight. While being a key component of ice protection system modeling, in practice, detailed surface water dynamics is difficult to observe. Improving the characterization of such icing test data has the potential to enhance physical understanding for engineers and contribute to the evolution of ice protection simulation models.

“Artificial intelligence holds promises of vast improvements in all areas of our industry: design, development, manufacturing, and operations. Bombardier is collaborating with AI experts in Canada and the UK to realize these promises and apply these emerging technologies toward the development of next-generation aircraft and rail products. With this challenge, we are creating opportunities for a potential collaboration that could evolve our ongoing research in icing dynamics and further improve the predictions of our simulation,” says Dr. Fassi Kafyeke, senior director of strategic technologies and innovation at Bombardier.

While aircraft operators use various heat- or material-related de-icing technologies to break or melt ice buildup during flight, better ice protection methodologies would mean improved aerodynamics and more economical fuel burn for aircraft that fly through extreme weather. These factors could cut costs for UK- and Canada-based operators.

In 2015, the UK government and industry committed to spend £3.9 billion ($50 billion) to further transform aerospace research until 2026 and build on technological strengths through the UK’s Industrial Strategy  . One of the strategy’s “grand challenges” is making the UK a global center for AI and data-driven innovation, which led the government to create the Artificial Intelligence Sector Deal  through the new Office for AI.

“This is a unique chance for the best and the brightest minds in artificial intelligence to come up with revolutionary ways to help aircraft operators to burn less fuel and cut costs. The UK has unrivalled heritage and world-leading expertise in both aerospace and AI, and our Industrial Strategy aims to build on that success in aerospace and make the UK a global center for AI and data-driven innovation,” says UK Business Secretary Greg Clark.

“Science and innovation has no borders, with most of the world’s best discoveries made through international and interdisciplinary collaborations. I have no doubt that working alongside Canadian researchers and innovators, UK experts will generate inspiring new approaches to this emerging technology,” continues Clark.

The AI Innovation Challenge will take open calls for participants through October 5 this year and will conclude with pitches delivered in UK and Canada on November 5 and December 3, respectively.

William Kucinski  is content editor at SAE International, Aerospace Products Group in Warrendale, Pa. Previously, he worked as a writer at the NASA Safety Center in Cleveland, Ohio and was responsible for writing the agency’s System Failure Case Studies. His interests include literally anything that has to do with space, past and present military aircraft, and propulsion technology.

Contact him regarding any article or collaboration ideas by e-mail at This email address is being protected from spambots. You need JavaScript enabled to view it..



Transcript

00:00:00 [Music] hi my name is Ben poal I'm the lead flight test engineer for FTV 2 which is just behind me uh we're here to talk about natural icing testing which we're currently carrying out on F32 uh it's the first time we carry out naturalizing testing on the C Series program it's a part of the Certification testing for every

00:00:29 airplane that's in service at the moment today so in preparation for the natural icing testing we took F32 to the paint shop and you can see the results in the black paint behind me uh we have two shades of the black paint there the matte black we're using that for the icing testing that allows us to see the white of the icing buildup and also any run backs that form when it melts and

00:00:50 then refreezes you can also see the gloss black paint on the Nels the tail and the uh above the Apu bake we'll be using that for the hot weather testing that allows us to leave the aircraft out get the maximum heating from solar radiation on the most critical components and verify that the aircraft's good to operate anywhere around the world no matter the

00:01:08 temperature additional painting you can see on the aircraft is the yellow and the red stripes and the yellow and black stripes in some locations uh we use these as ice witness um marks they're painted at 1 in intervals and allow us to estimate how thick the ice is formed and if it's forming in the correct locations and they've worked well on previous flights so far we've been

00:01:27 seeing between 3 and 4 in of ice at the worst case and it's matching well with our predictions from what we expect so far so earlier in the program we did some initial artificial ice shapes testing in that case we put the uh styrofoam shapes on the wings that simulate what we expect the ice Builder to look like and that's purely a handling test to ensure the aircraft

00:01:45 handles correctly with that ice on this is more of a systems test we go into clouds we allow ice to form naturally we see that the anti- ice system can stop it from accreting that it accretes in the shapes that we expected for artificialized shapes and also there's no differences to the handling qualities the aircraft with the Natural Ice as opposed to the artificial ice shapes so

00:02:03 in order to find the natural ice conditions uh we take advantage of a weather forecasting service from Ben Burn scene of Leading Edge Aviation he operates out of Colorado helps us with our briefs and also inlight through our sa phone Communications and he guides us to the best conditions of icing and our operating areas quite often around the Great Lakes um between 5 and 15,000 ft

00:02:23 in clouds we're looking for different types of icing so we look at droplet size water content and there's particular conditions that we want to see called the appendix C and we need to experience a good range of those in order to certify the aircraft for use in service so one of the special features we have an sv2 for the icing Tri is What's called the cloud combination

00:02:40 probe and that's the gold probe you can see on the aircraft behind me this consists of four instruments which measure liquid water content and droplet size and we use them to verify where we are within the appendix C uh we use a variety of combinations of the instruments depending on what we're seeing which have the best uh accuracy at that time and they measure the

00:02:59 droplet size and liquid water content with a variety of instruments including uh lasers Optical methods and what's called a hot wire [Music] prob so one of the difference with with the naturalizing testing is that we have a much larger crew than normal it depends a lot on the observations we make so from the normal crew of four we

00:03:23 operate on a day-to-day basis we're actually operating with a crew of seven so two pilots three flight test engineers and two of our IST from the work packs who are look at icing and make sure we get the right data we need so results are going well and we'll be continuing with the icing testing we'll be operating all around the northeastern of the US and Canada and you may see us

00:03:43 overhead anytime soon [Music]