Q2B 2020 | Quantum Computing Leadership | Grazia Vittadini, Marc Fisher | Airbus

Thanks a lot, Grazia. A very warm welcome also from my side. I'm very happy to be back at Q2B even if we can only meet virtually this year due to the pandemic. Flight physics is the broad denomination of all scientific and engineering aspects related to the flight of an aircraft. It encompasses a broad range of disciplines from overall aircraft design including aerodynamics, loads and aeroelasticity, mass properties, weight management, aircraft performance, and handling qualities. The future of aircraft design is very much dependent on finding ways of exploring new ways of optimization and multi-disciplinary optimization in particular. It is about new technologies and how they contribute to improve aircraft design and performance. The data and physics that govern the analysis of such technologies are complex and computationally demanding, particularly as many of the associated parameters are correlated across a framework of multi-disciplinary models or models of models. Addressing such complex systems requires altogether new capabilities and more powerful computational means for design space exploration and MBSE-enabled MDO. With classical computers gradually approaching their limit, quantum computing could potentially offer real breakthrough. Admittedly, the translation of our simulation and design problems into a quantum logic is non-trivial and might not even be feasible in every respect. As such, we believe that while for some problems full quantum solutions might be of interest, others are more likely to require a hybrid approach utilizing both quantum and classical computing capabilities. Often I am being asked the question: Why did you launch the Airbus Quantum Computing Challenge? First of all, we recognize that quantum computing expertise is not our core competency. However, we are very much interested and keen to gain an understanding of the maturity of quantum computing and then how it can be used to address aerospace-specific challenges, bringing science out of the lab into our aeronautical industry. Furthermore, we want to be part of this journey and stay true to our Airbus values in pioneering innovation. We are keen to understand which hardware is optimal for different types of problem statements. Therefore, as an end-to-end user, you could say that we remain hardware agnostic with a focus on which combination of approach and hardware provides the best solution for different problems at hand. The problem statements for this Quantum Computing Challenge were written with a balance between realistic and practical aerospace computational challenges in mind. We have deliberately chosen problems of different natures, some of which could actually be solved with conventional computational means. The point was to explore how problems of varying complexity would potentially translate into a quantum logic and benefit from quantum computing. As a recap, the five problem statements were: Problem statement number one focused primarily on aircraft climb optimization. An aircraft follows several flight phases during its trajectory from takeoff to landing. Cruise is the longest segment and is considered most important from a fuel and time optimization perspective. Yet for the ever-increasing volume of short-haul flights, climb and descend are more critical. Fuel optimization during these segments is very valuable for airlines. The challenge focused on the climb and how quantum computing can be applied to arrive at a low cost index which is central to climate efficiency. Problem statement number two featured computational fluid dynamics. The efficiency of aircraft design relies heavily on the aircraft's overall aerodynamic shape. The design is performed using computational fluid dynamics, so-called CFD, which is a both resource and very time-consuming task. This problem statement aims to show how established CFD simulations can be run using a quantum computer algorithm or alternatively be done in a hybrid way combining quantum computing with classical computing means. There were two objectives to this problem: finding a faster way to solve the problems and assessing how the algorithm scales in line with the problem complexity. Problem statement number three is about quantum neural networks for solving partial differential equations. Solving differential equations (PDEs) is a major challenge when solving aerodynamic problems. Today, their resolution requires complex numerical schemes and high computational costs. Traditionally, PDEs were solved in a deterministic manner using numerical methods. Recently, neural networks deep learning based algorithms have been deployed to solve coupled PDEs. These networks compute the time and space derivatives of a PDE, and the proposed challenge is to augment this new approach for aerodynamic problems with quantum capabilities. Problem statement four looked at wingbox design optimization. Given the limitations of classical computing, the aerospace industry faces a challenge in optimizing multi-disciplinary design, which is when design configurations such as airframe loads, mass modeling, and structural analysis must be simultaneously calculated. Quantum computing offers an alternative path to explore a wider design space by evaluating different parameters simultaneously, preserving structural integrity while optimizing weight. This balance is particularly important in aircraft wingbox design, where weight optimization is key in order to reduce operating costs and environmental impact. And last but not least, problem statement number five featured aircraft loading optimization. Airlines try to make best use of an aircraft's payload capability to maximize revenue, optimize fuel burn, and lower overall operating costs. Their scope for optimization is limited by the aircraft's operational envelope, which is determined by each mission's maximum payload capacity, the aircraft's center of gravity, and its fuselage shear limits. The objective of this challenge was to calculate the optimal aircraft configuration under coupled operational constraints. As mentioned by Grazia, we are really excited by the interest the AQCC has generated and the number of excellent submissions. After an exciting journey since the launch of the challenge some time ago and the announcement of the five finalists in October, we have now reached an important milestone in the challenge that will help us to shape our future computing landscape. And now it's time to go back to Grazia for the announcement of the AQCC winners.

First of all, thank you Grazia, thank you Mark, but also thank you to Airbus for this selection as winning team, but also for this international competition because I really think it was a great idea. Our company Reply is a consultancy firm and we want to stay ahead of the curve when it comes to technology and innovation. In fact, we have been working on quantum for the past three years. And when we saw your challenge, we just took it. The big question that everyone has in this field is: How do we map business problem to quantum computers? Of course, it's not an easy process, but being a consultancy firm and working in strict collaboration with business, that's what we're good at doing. But I also want to talk about the creativity of this process, because it's not only about technological mastery. For some new problems, you can just open up a book and read how to do it. For example, in the loading optimization problem, the difficult part were the constraints because we had respect constraints regarding the center of gravity or the distribution of the weights, and we had to find some new mathematical tricks to map all these different constraints into the same quantum computing model. But that was also the most fascinating part of this process, because at this core it was about innovation, which is the combination of creativity and technology. And it's great to see companies like Airbus that are investing in innovation. We are very happy and proud of this result. It really gives value to the work that we have been doing. And one personal thing I want to add is that for me being here today is amazing because three years ago I knew very little about quantum, but I was studying and reading quantum computing books of the same professor that are speaking at this conference in these days. So looking back, it really feels like an amazing journey. And the cool thing is that we are just getting started because of course we are looking forward to this collaboration in 2021, and I think it will be an amazing opportunity. So thank you again so much. I'll give it back to you, Graz and Mark. Well done, Machine Learning Reply.