Can cars or bikes drive themselves? Can ships or vessels steer themselves? Are drones or airplanes able to fly themselves? In this blog post, we explore how artificial intelligence is solving these questions for the aerospace industry.

For decades, commercial automotive companies built software to assist drivers in getting from one location to another. Driving assistance technologies such as emergency brake assists (EBAs) helped drivers deal with sudden breaking scenarios. Global positioning systems (GPSs) helped drivers navigate the streets with step-by-step driving directions.

For decades, commercial aerospace companies built systems to assist pilots in transporting passengers from one destination to another. Flight assistance technologies like radio communication, air traffic control (ATC) towers, flight manuals, and electronic flight instrument systems (EFISs) assisted pilots in navigating cloudy, stormy and low-visibility conditions.

While assistance navigational technologies were rapidly improving, engineers and programmers focused on automation of software and hardware. As a result, unmanned vehicles and autonomous machinery were born.

The development of unmanned vehicles began in the 21st century. During wartime, the aerospace companies were the first to develop autonomous aerial systems for the defense industry. Aerospace technologies quickly improved as funding from wars poured into the development of autonomous systems.

While the human conflict was the first catalyst in the development of unmanned vehicles, artificial intelligence (AI) was the next and final catalyst that allowed humanity to achieve autonomous navigation.

The problem with automation was that it was unable to handle the complexity of real-world scenarios and the dynamic nature of change. Engineers and software developers wondered if machines can be taught to imitate human behavior related to control and navigation. They asked themselves if “can machines be fully autonomous?” and more importantly “can AI understand the complexity of movement?”

One particular branch of AI — deep learning — turned out to be the answer to these questions. In this publication, we dive into the applications of AI for the aerospace industry.

Aerospace Industry

The aerospace industry is highly competitive and consists of many stakeholders in government, corporations, and defense agencies. Globally, aerospace is worth $838 Billion USD according to AeroDynamic Advisory and Teal Group Corp [1]. The aerospace industry constitutes all activities relating to the development, production, maintenance, and support of aircraft, spacecraft, and rotorcraft. We wrote this article as a guide to anyone involved in project development, component manufacturing, construction, operations, and education, training, and research within the aerospace industry.

Artificial Intelligence (AI)

AI and deep learning is the catalyst for autonomous vehicles. Engineers combined visual data from cameras with neural networks to develop autonomous AI systems without the reliance of human control or navigation. Autonomous vehicles leveraged 3D object detection and semantic segmentation to travel on roads by themselves. Autonomous drones used image classification to follow a path without reliance on global positioning systems (GPSs). Autonomous aerial vehicles leveraged object classification and object detection in aerial images to navigate from the bird’s-eye view perspective.

AI Research in Aerospace Industry

Artificial Intelligence research and academia are often ahead of the industry. Large budgets are dedicated to research and develop state-of-the-art methods in machine learning and deep learning algorithms. Reading AI research is time-consuming and the process requires an understanding of technology and computer science terminology. Luckily, the Produvia team spent dozens of hours reading the research papers, so you don’t have to.

Design and Product Development

The design of aerial vehicles requires creativity. Product development in aerospace is necessary to develop new solutions. Luckily, AI is able to assist in both [2–4].

  • Knowledge Transfer in Aircraft Design and Product Development
  • Estimation of Manufacturing Cost of Jet Engine Components

Personal Health

The health of pilots is critical. If pilots are unable to regular their stress or experience health conditions, they risk the operation and aerial machinery [5].

  • Heart Attack Prediction
  • Stress Prediction

Structural Health

Each aerial machinery must be maintained and inspected regularly. AI can help answer questions related to the structural health of hardware [6–13]

  • Rare Failure Prediction
  • Diagnosis of Aerospace Structure Defects or Damages
  • Damage Detection of Impact
  • Structural Health Monitoring
  • Anomaly Detection and Fault Diagnosis for Space Systems
  • Diagnosis of Rotorcraft
  • Damage Location Detection
  • Predictive Maintenance of Electrical, Mechanical, or Structural Systems
  • Predictive Maintenance using Drones or Unmanned Aerial Vehicles (UAVs)

Flight

Autonomous flight is the holy grail of aerospace. AI and deep learning is helping companies accomplish this complex goal [14–18]

  • Horizon Detection
  • Autonomous Drones
  • Autonomous Helicopters
  • Autonomous Aerospace Vehicles

AI Applications in Aerospace Industry

There are many companies and organizations that already incorporate AI, machine learning, and deep learning into their products and services. Here are our favorite AI applications in the aerospace industry:

  • Uptake offers artificial intelligence solutions for failure prediction, noise filtering, image analytics, anomaly detection, recommendations, data integrity, dynamic rules, fuel management [19]
  • SparkCognition uses machine learning to warn of aircraft and asset failures before they occur, maximizing fleet availability, minimizing unscheduled maintenance, and extending asset life. SparkCognition also uses natural language processing to reduce troubleshooting time by automatically classifying fault codes and recommending the best corrective actions. SparkCognition uses reinforcement learning to provide a simulated environment in which a user can train control algorithms and evolve swarm tactics. [20]
  • Slingshot Aerospace helps aerospace companies to manage risks and threats, and detect and map debris using predictive analytics, geospatial signal processing, and machine vision. [21]
  • Neurala uses computer vision and deep learning technologies to classify images and to recognize objects to avoid obstructions autonomously [22]
  • Orbital Sidekick leverages machine learning for camouflage detection, plume detection, target identification, material classification, and space situational awareness [23]
  • Boeing features several autonomous systems, from the seabed to space, including autonomous maritime vehicles (autonomous surface platforms, unmanned undersea vehicles), unmanned aerial systems, and autonomous space vehicles. [24]

AI Ideas for Aerospace Industry

There are many artificial intelligence ideas to explore. AI R&D projects are greenfield projects. There are many opportunities, but coming up with ideas can be challenging. That’s why I brainstormed a few AI ideas to help you get started in the aerospace industry:

  • Satellite Image Analysis. It’s possible to develop deep learning models that aim to automate image analysis of satellite data. This can be accomplished by solving object detection, semantic segmentation, change detection, and super-resolution tasks.
  • Fault Detection and Failure Detection. Deep learning models can be developed to detect faults and failures in complex aerospace systems. This can be accomplished by solving anomaly detection and recommendation system tasks.
  • Autonomous Exploration. Machine learning can be used in planetary exploration rovers. By solving calibration, visual odometry, robot navigation, motion planning, robotic grasping, safe exploration, robot task planning, autonomous vehicles, and autonomous navigation tasks.
  • Power or Fuel Consumption Prediction. Machine learning can be used in spacecraft engineering to predict the power or fuel consumption of the spacecraft.

About Produvia

We help companies implement artificial intelligence technologies. Interested in exploring these or similar AI use cases in aerospace? Visit our website to book a discovery call with the Produvia team.

References

  1. Aerospace Industry Is Worth US$838 Billion. (2019). Aviationweek.com. Retrieved 26 August 2019, from https://aviationweek.com/farnborough-airshow-2018/aerospace-industry-worth-us838-billion
  2. Knowledge Transfer Through Machine Learning in Aircraft Design — IEEE Journals & Magazine. (2019). Ieeexplore.ieee.org. Retrieved 26 August 2019, from https://ieeexplore.ieee.org/abstract/document/8065136
  3. Bertoni, A., Dasari, S., Hallstedt, S., & Petter, A. (2018). Model-based decision support for value and sustainability assessment : Applying machine learning in aerospace product development. The Design Society, 6, 2585–2596. Retrieved from http://www.diva-portal.org/smash/record.jsf?pid=diva2%3A1210597&dswid=6877
  4. Loyer, J., Henriques, E., Fontul, M., & Wiseall, S. (2016). Comparison of Machine Learning methods applied to the estimation of manufacturing cost of jet engine components. International Journal Of Production Economics, 178, 109–119. doi:10.1016/j.ijpe.2016.05.006
  5. Decision making system using machine learning and Pearson for heart attack — IEEE Conference Publication. (2019). Ieeexplore.ieee.org. Retrieved 26 August 2019, from https://ieeexplore.ieee.org/abstract/document/8212797
  6. Burnaev, E. (2019). Rare Failure Prediction via Event Matching for Aerospace Applications. arXiv.org. Retrieved 26 August 2019, from https://arxiv.org/abs/1905.11586
  7. D’Angelo, G., & Rampone, S. (2014). Diagnosis of aerospace structure defects by a HPC implemented soft computing algorithm. 2014 IEEE Metrology For Aerospace (Metroaerospace). doi:10.1109/metroaerospace.2014.6865959
  8. D’Angelo, G., & Rampone, S. (2016). Feature extraction and soft computing methods for aerospace structure defect classification. Measurement, 85, 192–209. doi:10.1016/j.measurement.2016.02.027
  9. Staszewski, W., Mahzan, S., & Traynor, R. (2009). Health monitoring of aerospace composite structures — Active and passive approach. Composites Science And Technology, 69(11–12), 1678–1685. doi:10.1016/j.compscitech.2008.09.034
  10. Structural Health Monitoring. (2019). Google Books. Retrieved 26 August 2019, from https://www.wiley.com/en-us/Structural+Health+Monitoring%3A+A+Machine+Learning+Perspective-p-9781119994336
  11. Telemetry-mining: a machine learning approach to anomaly detection and fault diagnosis for space systems — IEEE Conference Publication. (2019). Ieeexplore.ieee.org. Retrieved 26 August 2019, from https://ieeexplore.ieee.org/abstract/document/1659593
  12. Wade, D., & Wilson, A. (2019). 17th Australian International Aerospace Congress : AIAC 2017 — Applying machine learning-based diagnostic functions to rotorcraft safety (Engineering Collection) — Informit. 17Th Australian International Aerospace Congress : AIAC 2017, 663. Retrieved from https://search.informit.com.au/documentSummary;dn=740994444756033;res=IELENG
  13. Fu, H., Vong, C., Wong, P., & Yang, Z. (2014). Fast detection of impact location using kernel extreme learning machine. Neural Computing And Applications, 27(1), 121–130. doi:10.1007/s00521–014–1568–2
  14. Horizon Detection Using Machine Learning Techniques — IEEE Conference Publication. (2019). Ieeexplore.ieee.org. Retrieved 26 August 2019, from https://ieeexplore.ieee.org/abstract/document/4041464
  15. Autonomous drones for assisting rescue services within the context of natural disasters — IEEE Conference Publication. (2019). Ieeexplore.ieee.org. Retrieved 26 August 2019, from https://ieeexplore.ieee.org/abstract/document/6929384
  16. Adaptive Trajectory Control for Autonomous Helicopters | Journal of Guidance, Control, and Dynamics. (2019). Journal Of Guidance, Control, And Dynamics. Retrieved from https://arc.aiaa.org/doi/abs/10.2514/1.6271
  17. Trajectory Planning for Autonomous Aerospace Vehicles amid Known Obstacles and Conflicts | Journal of Guidance, Control, and Dynamics. (2019). Journal Of Guidance, Control, And Dynamics. Retrieved from https://arc.aiaa.org/doi/abs/10.2514/1.12514?journalCode=jgcd
  18. Real-Time Trajectory Planning for Autonomous Aerospace Vehicles amidst Static Obstacles | Infotech@Aerospace Conferences. (2019). Arc.aiaa.org. Retrieved 26 August 2019, from https://arc.aiaa.org/doi/abs/10.2514/6.2002-3421
  19. Industrial AI and IoT Platform | Data Science for Global Industry. (2019). Uptake. Retrieved 12 September 2019, from https://www.uptake.com/platform
  20. AI in Aviation | Artificial Intelligence in Aviation Industry. (2019). SparkCognition, Inc.. Retrieved 12 September 2019, from https://www.sparkcognition.com/industry/aviation/
  21. Slingshot’s Data Platform — Slingshot Aerospace — Next Generation Signal Processing AI. (2019). Slingshotaerospace.com. Retrieved 13 September 2019, from https://www.slingshotaerospace.com/products/slingshots-data-platform-clairvoyance/
  22. Brain Builder. (2019). Neurala.com. Retrieved 12 September 2019, from https://www.neurala.com/solutions/brain
  23. Boeing: Autonomous Systems. (2019). Boeing.com. Retrieved 13 September 2019, from https://www.boeing.com/defense/autonomous-systems/index.page
  24. Markets | Orbital Sidekick. (2019). Orbital Sidekick. Retrieved 12 September 2019, from https://orbitalsidekick.com/markets/#defense