Aiir Innovations: taking AI to the skies

Aiir Innovations: humble beginnings, big impact

“The most beautiful part of the whole thing is when we started the project we didn’t really know each other, but slowly we’ve grown up together.” It’s a remark that you might expect to hear in a scene from one of your favourite movies, or perhaps in a confession from a colleague on a team-building day. But it’s actually how AI tech startup Aiir Innovation’s CEO Bart Vredebregt explains how 5 AI students and an AI professor came together to launch a business that could revolutionise the maintenance industry.

It all started in 2015, when Dutch airline KLM invited students on the University of Amsterdam AI master’s degree programme to work on a project to automate the inspection of their jet engines. When done manually, this inspection requires an engineer to put a camera into the engine after removing the aircraft's wing and can take up to 24 hours to complete.

“As the inspection can take so long engineers can get tired and start making mistakes. The team at KLM wondered if we could help automate this process, which we showed with a proof-of-concept that we could. After we had shown them it could be done they asked us where they could get the technology from, and as it wasn’t available elsewhere we launched the company to help provide it to them.” Thus Aiir Innovations was born. 

Specialist knowledge of AI and machine learning

5 members of the team studied for a master’s degree in AI, and all have individual specialisms including research, computer vision and data analytics. The team member is the person that taught them during those studies, Professor Jan van Gemert, who is head of the computer vision group at Delft University of Technology.

Despite their excellence in AI, machine learning and computer vision, the group quickly learned that launching a startup wasn’t just about programming. Luckily, they were introduced to the team at the Amsterdam Center for Entrepreneurship (ACE), creators of Amsterdam Science Park’s Startup Village, who helped teach them how to be entrepreneurs.

“After starting the company in 2016 we spent that first year learning how to run a business, and then slowly started to build the system we have now. Not one of us has an MBA or business background, so ACE helped us learn all about business. They also helped us secure our premises at the Startup Village – we were actually the first firm to move in. Getting our own office there was when we really felt like entrepreneurs for the first time.”

The perfect base at Amsterdam Science Park

Being based at Amsterdam’s ground-breaking Science Park has helped Aiir recruit talent and share knowledge with others in their sector. “It’s a great base for what we do, as having that address helps attract people who know its reputation, both in Amsterdam and internationally. People sometimes underestimate how much work is being done around AI in Amsterdam, including great research at the universities and big successes in business, such as companies like Scyfer. Amsterdam's universities also offer amazing courses in AI.”

Revolutionising inspections in the aviation industry

Aiir now has a state-of-the-art anomaly detection system that uses large sets of data to learn what a normal engine should look like during a video inspection. It then uses that data to detect deviations from the normal standard and alert engineers to potential problems and faults. The system currently comes in one package with both the hardware and software housed in a suitcase that looks straight out of a James Bond film. Vredebregt likens Aiir’s rapidly improving system to that of an email spam filter, which learns what should and shouldn’t be contained within an inbox. 

“We have made something which assists the engineer during the inspection, rather than being fully automated. It’s like an extra pair of digital eyes. After doing some tests we are now aiming to launch the system fully by the end of the year with KLM. This will bring AI into the real world and show that it’s possible to automate inspections like this, then we aim to market the system to airlines around the world by the end of next year. What’s great is that the team at KLM have been right behind us in supporting us to make a product we can sell to others, rather than trying to keep it to themselves.”

Developing the technology of the future

Aiir is currently funding itself by carrying out private AI consultancy work for various clients while focusing on refining its system for use in the aviation sector. But Vredebregt says that eventually its technology could be used for a wide variety of applications using video footage from any industry.

“We think it will be easier to pursue that when we can show we have a track record in the aviation industry, but our system could be used for so many different purposes to list them would take too long. Humans make mistakes if they get distracted or tired, that’s where AI can help.”

Vredebregt says predicting the next big trends in AI is difficult, as the technology’s nature makes its development unpredictable. For Aiir Innovations, the future involves continuing their hard work to make their system work as efficiently as possible, and, of course, supporting each other in making the business the best it can be.

“I think it’s a good thing that we can’t predict what will be the next big thing, as there’s so much cool stuff going on in AI. When I think about where we have come in just a few years it’s crazy. A lot of people warned us that 6 people was a big number of people to launch a startup, but we thought ‘let’s just do it’. Now we have a really strong bond and work so well together. Everyone has grown a lot, in terms of their technical skills, and their social and business skills. I am really looking forward to seeing where everyone will be in another 3 years, because if we keep improving like we have done, then the sky's the limit.”