NATS – The challenges of absolutely automating air site visitors administration – Drones Information
This week, I spoke about the challenges of fully automating ATMs at the event organized by the Aerospace Technology Institute, “The Journey to Autonomy in Civil Aerospace”.
We tend to think of autonomous things as self-driving cars or machines that do things without human input, but automation is something that we get used to in our daily lives today. From our phones suggesting for us to our televisions automatically recording shows we know we’ve seen before.
Automation for the aviation industry offers tremendous opportunities and has the potential to open the skies to new airspace users and allow us to be more flexible and agile in our services. It also brings with it some challenges, and there are three that I think need to be considered before anyone else.
The first is security. Security for NATS is what we do every other day. Our job is to move planes safely from one location to another as efficiently as possible. The safety of thousands of flights carrying hundreds of thousands of people every day is up to us. To ensure this security, layers of mechanisms and procedures are embedded in our work.
Automation can bring with it the possibility of further improving the level of security. And it’s already in operation. The big leap for the future will come from the controllers who make the decisions with tools to support them. on the technology that makes the decision without a person checking and then “accepting” the solution.
We often compare the automation of the aviation industry with that of autonomous cars, but in reality the safety levels within the two industries are not comparable. We need even stricter acceptance criteria. In the UK alone, around 27,000 people are killed or seriously injured in car accidents each year – the same number of people it would take to fill 180 Airbus A320s. In 2019 there were approx. 257 commercial aviation deaths worldwide. The level of security required to implement an automation must reflect this additional level of security.
This leads to the second challenge: complexity. Airspace is complex and the way we manage it requires skill and judgment. The training to become an air traffic controller takes around three years after a tax selection process. The reason why the human brain is so good at solving problems in this environment is that it can process a lot of information, but most importantly it can also deal with ambiguities. A machine can handle a lot more information, but no ambiguity. How do we make sure it can handle a new scenario that it has never seen before? How does a machine ensure that the answer it generates is safe and efficient? It has to be 100% correct.
Another complexity is our neighbors. We are working with European partners to harmonize air traffic management. However, if the UK had a fully autonomous ATM system and our neighboring ANSPs didn’t, it would make the interface more than difficult.
The third challenge is human acceptance of automation – whether it’s traveling passengers, pilots, or regulators. If man doesn’t trust technology, it may never reach its full potential. Passenger adoption is important, but as we advance the automation journey in the ATM environment, trust between controllers and technology is essential and therefore they are an integral part of the evolution of that technology and systems.
The success of COVID-19 has really shown how external factors affect the industry and how we need to remain adaptable and flexible. A technical solution today can be out of date in a few years. However, we know that automation is playing an increasing role in helping our air traffic controllers provide the safest, most efficient service for aircraft flying through our airspace.