Is Boeing Switching to Autopilot? – Machine Learning in the Aerospace Industry

Machine learning is perceived to be the holy grail of business rejuvenation and its adoption a direct route to capturing future markets. Will the aerospace industry and Boeing prove that these expectations may not always be straight out of “la la land”?

0.01%. This is probably close to the percentage of business managers who understand what machine learning (ML) really is and what it can do for a business. This is no surprise considering that this figure is also fairly close to the percentage of professionals who understand it – including engineers developing ML (Kesmodel 2018). This doesn’t seem to stop people from touting AI and ML as the geese that will inevitably lay golden eggs for businesses in the near future (Loten 2018). People in all industries – customer service, energy, health, or essentially any kind of manufacturing – expect miracles from the use of ML. There might well be one industry however, where ML will soon bring immense gains and a company which could soon demonstrate this.

Manufacturing in the aerospace industry exhibits a set of unique characteristics which make it ideal for learning based process automation. And it is in desperate need of it. Boeing knows this and is investing heavily into the technology (“Boeing’S Venture Arm Invests In Artificial Intelligence, Machine Learning Company – Avionics” 2018). Perhaps it is partly motivated by looking towards Europe. Although Airbus employees get a good chuckle out of referring to the engineless A320neo airliners piling up in Toulouse as their “glider-range”, the engine delivery problem caused by supply-chain management errors is no laughing matter. The announcement caused a near 5% drop in share prices a few months ago (Kotoky and Katz 2018). This is a key type of problem which the use of ML could potentially prevent.

Machine learning is well suited to improve self-contained processes involving complex statistical relationships (Fedyk 2016) such as supply chain management challenges, maintenance prediction or the fast restructuring of assembly lines based on demand changes. These are the exact areas Boeing is looking at utilising ML in the short and long term (Rao 2018).

Predictive maintenance uses AI and big data (a constant stream of copious data) on failure symptoms, causes and previously used solutions, to build a model capable of both diagnosing and forecasting failures. Natural language processing – a key function of current AI technologies – enables the fast absorption of manuals and relevant documentation (repair notes, etc.) and build a predictive maintenance model. This is one of the systems which SparkCognition – a company Boeing has invested in then partnered with (Garcia 2018) – has developed. The collaboration with SparkCognition was originally overseen by Boeing’s venture arm, HorizonX. This branch focuses on investment into future technologies and they are clearly looking in the right directions. Customers report that Boeing’s predictive maintenance services to airlines lead to an 80% reduction on maintenance issues (Canaday 2018).

Long term goals in using ML can also be seen in Boeing’s manifesto for future developments. One key focus is the use of AI supported supply-chain management systems. Boeing is clearly looking at ways it can avoid blunders like the glider-fleet sitting at Airbus’ Toulouse site. The other direction is ways to revolutionise factories.

The company’s final-assembly lines in Renton are capable of churning out as many jets as Airbus can in three separate factories combined (Gates 2018). Investments here previously focused on re-configuring the assembly lines adopting a Kaizen-style lean approach (Kesmodel 2018) but have recently moved on to full automation using robots (Grunbaum 2018). For now without the large scale use of AI. Certain applications for ML in manufacturing – such as in the choice of metals – are already present however (Chen et al. 2018).

Large scale application of ML in manufacturing is already planned at Boeing and the company is approaching it the right way. One of the pitfalls of the use of ML is that it can require a high degree of customisation and can lead to problems such as a more rigid plant configuration – potentially disastrous in a fast changing market. This is what Mercedes Benz (Wilson and Daugherty 2018) and in fact Airbus (Gates 2018) has also recognised. Flexibility requires the use of human-machine systems – cobots. Cobots or collaborative robots are created to aid human activities in a plant. They are also easier to set up to be quickly reconfigured by their human counterparts to support different operations.

The reality is that we are very far from Ray Kurzweil’s singularity (Kurzweil 2006) and current applications of AI are often motivated by solutionism. Many question how far AI can go in the future (Somers 2018) and some say that its use may simply end up re-creating challenges at a different level (Willcocks 2016). However, as strong growth seems to persist in the airline market and it continues to move at breakneck speed, it seems the fast adoption the of this new technology could be crucial to future success. If the use of ML can yield gains in any sector, it will be the aerospace industry. Boeing seems to be conscious of the risks of using ML and is approaching it in the correct way.

Whether ML will truly live up to the unquestionably high expectations and if Boeing can maintain a smart approach, resisting the solutionist and blind adaptation is yet to be seen.

(Word count: 793)



Kesmodel, David. 2018. “Boeing Teams Speed Up 737 Output”. WSJ.

Loten, Angus. 2018. “Tech Chiefs Need To Manage AI Expectations”. WSJ.

“Boeing’S Venture Arm Invests In Artificial Intelligence, Machine Learning Company – Avionics”. 2018. Avionics.

Kotoky, Anurag, and Benjamin D Katz. 2018. “Airbus Will Miss Its A320neo Delivery Goal After Engine Problems”. Bloomberg.Com.

Fedyk, Anastassia. 2016. “How To Tell If Machine Learning Can Solve Your Business Problem”. Harvard Business Review, , 2016.

Rao, Harish. 2018. “Boeing: AI Driven Transformation”. Boeing.Com.

Garcia, Marisa. 2018. “Boeing Partners With Sparkcognition To Develop AI Solutions For Future Of Air Transport”. Forbes.Com.

Canaday, Henry. 2018. “Continued Progress Under Boeing’S Predictive Maintenance Umbrella”. MRO Network.

Gates, Dominic. 2018. “Boeing Retools Renton Plant With Automation For 737’S Big Ramp-Up”. The Seattle Times.

Kesmodel, David. 2018. “Boeing Teams Speed Up 737 Output”. WSJ.

Grunbaum, Rami. 2018. “Boeing Video Showcases Robots That Will Help Build 777”. The Seattle Times.

Chen, Sophia, Sophia Chen, Katie Palmer, Matt Simon, Rhett Allain, Shaun Raviv, Matt Simon, Megan Molteni, and Matt Simon. 2018. “The AI Company That Helps Boeing Cook New Metals For Jets”. WIRED.

Gates, Dominic. 2018. “Airbus, Like Boeing, Pushes To New Heights Of Automation”. The Seattle Times.

Kurzweil, Ray. 2006. The Singularity Is Near. New York: Penguin Books.

Somers, James. 2018. “Progress In AI Seems Like It’S Accelerating, But Here’S Why It Could Be Plateauing”. MIT Technology Review.

Willcocks, Leslie P. 2016. Service Automation: Robots And The Future Of Work 2016. Steve Brookes Publishing, 2016.



Crowdsourcing legacy: How Rio is changing policymaking


Smart from start to finish: GM’s path to new processes and products

Leave a comment