With demand for air travel escalating, it’s no secret that airlines and airports will require more robust technology to ensure that aircraft can arrive, pick up passengers, and depart as swiftly as possible.
Zurich and US-based tech specialist Assaia has capitalised on artificial intelligence (AI) technology in a bid to minimise delays and optimise airport turnarounds. The company’s Assaia Apron AI solution analyses footage from existing cameras surveying the airport apron, facilitating real-time predictive analysis that can help airlines identify the reasons why aircraft are running late and implement effective prevention measures.
With clients such as British Airways, Gatwick Airport, Swiss Port, EuroAirport and Toronto Pearson among others, Assaia claims its technology will help prevent accidents, ensure efficient usage of runways and boost revenue. To find out more, we spoke to Assaia chief customer officer Christiaan Hen.
Varsha Saraogi: How did Assaia founders Nikolay Kobyshev and Max M. Diez come up with the Apron AI solution?
Christiaan Hen: They are highly skilled in computer vision and AI and they knew they had something valuable but didn’t know what to do with it. They started organising workshops for companies to showcase the technology [and] to find a business application.
In one of those workshops the head of innovation from Swissport was present and when he saw this he said: ‘If this actually worked, and you can produce real-time videos and insight about what is happening in an internal process of an aircraft then it would be super valuable for everyone in the industry.’
So then they started getting talking with airports and airline companies to show how the technology actually worked. There was so much enthusiasm on this application that within two months or so they decided to fully zoom in on the aviation sector and said ‘we are now an aviation company’.
VS: How does the technology work exactly?
CH: The basic functionality of the system is that we take video screens from an existing cameras already installed at the airports and we turn that video data into structured data. Structured data includes timestamps – when does the catering truck arrive and when [does it] it depart, for instance. Then airline companies can use the information through these recordings and do an analysis about turnarounds, [find] the root causes of delays, and they can prevent those problems in the future.
For example, if I know when an aircraft should be arriving and if it doesn’t arrive I can immediately act upon that, rather than having only five to ten minutes before someone actually notices a plane is not there – which is a lot of time if you consider the 30 or 40 minute turnaround.
Another use is to predict the off-block time of the aircraft. These predictions are more accurate than the current data alternative which is the Target Off-Block Time from the A-CDM framework and if you know in advance which aircrafts are going to be delayed, you can add additional resources to prevent this delay from happening.
VS: What are the main challenges facing airline companies which Assaia intends to resolve?
CH: There’s three main areas where the value is. One is increased on-time performance, second is increase safety and the third one is decreasing operational costs.
If we reduce delays, then we can reduce costs associated to delays. Additionally, you can optimise your turnarounds which means you could increase aircraft utilisation, and that is going to add revenue for the airlines. You can also use this information to optimise gate planning, push back sequencing, runway sequencing as well as track the speed of objects.
You could compare this to the Formula One pit stop. There’s a lot of guys doing different things – one is changing tires, one is cleaning helmets, another guy is fuelling the car and a guy is even changing parts. But it is super organised and synchronised because everyone knows exactly what to do. And you can turn around the car in under two minutes. If you look at the aviation industry, it’s very different because it’s uncoordinated. Everyone is working on schedules that have already been overtaken by unexpected incidences and therefore they can’t manage to turn around planes as fast as they should.
If you look at airports in general, they need smart tech-based solutions. I worked for Amsterdam Airport for over eight years as the head of innovation. One of the main reasons that a lot of money was being poured into the digital programme was because we knew we had to use smart solutions because otherwise we just wouldn’t be able to handle more traffic. So it’s not something we made up, but it’s a trend we see in the industry. Since our product is video-based it’s easy to show it to people and they quite quickly get it. And that is also one of the reasons why we’ve been able to grow very fast since March 2018.
VS: What was the reason to specifically pick the aviation industry to use this technology?
CH: Forecast says the demand for air travel is going to double in the next 20 years and if we don’t come up with smart solutions then we will just not be able to accommodate this growth. Airports [are] in the infrastructure sector so whenever there’s a problem, the initial solution would be adding infrastructure.
The physical space in a lot of the major airports is limited. And adding infrastructure requires time and money. I’m not saying we shouldn’t do that, but it’s not going to be enough.
VS: Using AI technology and machine learning can be expensive. Is that a challenge for airports?
CH: As far as Assaia is concerned, if we cannot create a plausible business case for our customers we will go out of business. I’m not saying that it’s cheap, but the potential value that it creates is huge. Every single time I do these calculations, I’m flabbergasted by the amount of money that could be generated if you put the technology to use.
If an airport is interested to make the most out of AI then they can get an insane return on investment.
VS: Where do you see the trend of automation going for airports in the near future?
CH: Our vision for airports is it will soon go towards full automation. You can already see airports experimenting with autonomous ground support equipment (GSE), for example. The interesting part is if you’re using an autonomous GSE, who’s going to tell these GSEs where and when to go? That’s where we position ourselves because we have full visibility on the internal runs in an airport. We could then act as an operating system for this autonomous GSE.
Things like decision-making based on a lot of variables [are] typically something that humans are not particularly good at and it’s something that AI is very good at. So we are using the camera as a sensor to generate data which can predict off stock time. You can use these predictions, for example, in data centre allocation. Over time, the systems will evolve in to becoming fully autonomous.
We were also looking into using the technology to validate visas and passports, and even in baggage handling.