Innovative solutions driving your business forward.
Discover our insights & resources.
Explore your career opportunities.
Learn more about Sogeti.
Start typing keywords to search the site. Press enter to submit.
Generative AI
Cloud
Testing
Artificial intelligence
Security
January 27, 2023
KLM is training its pilots partly in virtual reality (VR). To this end, the airline organization has developed a digital cockpit together with Sogeti and others: a virtual environment that resembles a real cockpit in every detail. With VR goggles on, it looks just as if a pilot is sitting in the cockpit. They can also inspect the plane inside and out. That makes the digital cockpit the ideal tool to largely train pilots in both common and rare and extreme situations.
A cockpit has lots of buttons and procedures. When one of them changes, the entire digital cockpit must change with it. Then it must be checked that everything in the VR application is still working properly in connection with the high safety standards in aviation. This is done through regression testing.
Has anything changed in the digital cockpit? KLM used to have to manually perform such a regression test for every button and procedure. A time-consuming task with the possibility of overlooking things in the test environment.
To prevent this and to test faster, KLM wanted to start automated testing of the digital cockpit. How can we do that efficiently and smartly? KLM put this question to a team of VR specialists, software developers and testers from Sogeti. During a brainstorming session, they came up with the unique solution: use the internal Artificial Intelligence (AI) functionality from the VR application to test the digital cockpit.
Full of enthusiasm, a team of Sogetists went to work. In a pressure cooking session, developers and engineers together arrived at a groundbreaking insight: why not have the regression tests performed by the AI functionality in the VR application itself? No sooner said than done.
Together with KLM, Sogeti first developed a Proof-of-Concept for the test automation of the digital cockpit. Here, regression tests are performed by ML Agent, the AI functionality in the VR application itself. ML Agent is part of the Unity platform on which the VR application is built. Previously, this AI tool was only used in the training itself, for example to simulate scenarios with other aircraft or unexpected weather conditions. Professionals from KLM and Sogeti trained the AI test functionality together.
Thanks to VR test automation, testing the digital cockpit is much less labor-intensive. The testing tool takes repetitive operations out of their hands. This leaves them time for work where human brainpower is required. Think of coaching or giving feedback. Moreover, regression tests are now performed much faster. The ML Agent has checked the operation of all buttons and procedures in seconds. On top of that, regression tests are extremely accurate, so any small error is removed from the application.
ML Agent’s Proof-of-Concept was a great success, which is why KLM is including this AI test automation in the new version of the digital cockpit currently under development. With this, KLM is also taking steps toward more sustainable flight training. Pilots in training can now do a large part of their training hours virtually, on the ground. The Virtual Cockpit and ML Agent thus not only ensure a higher quality of training, but also enormous time savings. The Virtual Cockpit and ML Agent thus provide, among other things, better quality training, more accessible way of frequent training, effective training of specific components and a time saving! Moreover, training can continue as usual during Covid-19 outbreaks.
Also better software quality through lightning-fast, error-free regression testing? Take a look directly at the expertise Sogeti can offer in the field of AI.
Find out more
Client: Air France KLM
Region: Netherlands
Industry: Aviation
Offer: Quality Engineering & Testing
1 PDF (950 KB)
We use cookies to improve your experience on our website. They help us to improve site performance, present you relevant advertising and enable you to share content in social media.
You may accept all cookies, or choose to manage them individually. You can change your settings at any time by clicking Cookie Settings available in the footer of every page.
For more information related to the cookies, please visit our cookie policy.