Automation vs. autonomy in urban air mobility

Today, self-piloting technologies are having a huge impact on mobility solutions—both on the ground and in the air. But the leap from automated to autonomous is still a work in progress. Although fully autonomous aircraft are still many years away, urban air mobility vehicles are proving to be an excellent option to start the roll-out of self-piloting aircraft operations.

Greater reliability, improved performance and reduced costs: the perceived benefits of automation have made it one of this century’s most significant industrial trends. In fact, automation in manufacturing has become so widespread that robots, artificial intelligence and machine learning are now an integral part of the manufacturing process. 

In mobility, automation is also set to make a significant impact on how we move in the future. Automotive manufacturers around the world are exploring the potential of self-driving cars, which promise to drastically reduce vehicle accidents, improve road safety and free up valuable time.  

While certain levels of automation are already available in many vehicles—such as cruise control, steering assistance and self-parking—it will be a number of years before self-driving cars hit the roads. This is, in part, because road systems are highly complex environments in which hazards abound—from other road users to pedestrians and debris.       

The sky, however, is a very different story. Aircraft operate in a much sparser environment. Air hazards—such as other aircraft, birds, and drones—are rarely encountered compared to ground hazards, which are often minimised by applying safety means at landing and take-off sites. If a hazard is encountered, the aircraft can move in three dimensions to avoid it.

When viewed from the perspective of risk and impact, this vast difference in hazard density and response options suggests that autonomy in aviation faces a much different challenge in getting off the ground than its automotive counterparts.

The leap from automated to autonomous

Although sometimes used interchangeably, automation and autonomy are not synonymous. The difference can be summarised as follows:

• Automation refers to the ability of a system to control a vehicle, like autopilot or cruise control.
• Autonomy is the ability of a system to not only control a vehicle but respond to unexpected hazards.

Two very different paths can be taken to achieve a fully autonomous aircraft:

• Fall-back pilot: During development, a “fall-back,” or safety pilot is always on board to take control at any time, for any reason. As the autonomy system becomes more reliable, the fall-back pilot will eventually become redundant.
• Full autonomy from the beginning: Operating solely in constrained environments enables the autonomy system to be implemented from the very beginning for safe and extensive testing

In ‘autonomous’ flight, the aircraft has to be able to make decisions and react to unforeseen events without the pilot’s intervention.

- Arne Stoschek, Project Wayfinder, A^3 by Airbus

Using the fall-back pilot enables the aircraft to become airborne more quickly, but it requires significant investment in systems that will, ultimately, not be needed. Starting with full autonomy eliminates the need for human-machine interfaces, but obtaining certification and public acceptance can be a challenge.

Some manufacturers, like Airbus, are already taking the latter option: new urban air mobility vehicles like Vahana have been designed to integrate self-piloting functionalities from the start. This is because flying taxi missions involve short, point-to-point flights along a restricted selection of routes using limited landing infrastructure, thus making urban air vehicles an ideal choice for testing self-piloting operations in aircraft. Although the technology is still not mature enough to transport passengers, autonomy in aircraft operations will undoubtedly be shaped by self-piloting, flying taxi demonstrators like Vahana.

3 questions with Arne Stoschek, autonomy expert at A3 by Airbus 

Why is it important to distinguish between automated and autonomous?

There’s often some confusion between the two, but as autonomy becomes increasingly important, we need to make sure we avoid this. “Automated” refers to a set of prescribed actions. In flight, these actions are triggered by a pilot, such as automated landing. So, the pilot initiates it and the aircraft then executes it. In “autonomous” flight, however, the aircraft has to be able to make decisions and react to unforeseen events without the pilot’s intervention.

If the objective of self-driving cars is to drastically reduce vehicle accidents and improve safety on the road, what is the objective of self-piloting air vehicles?

Self-piloting technology in aircraft addresses two main topics. One is the predicted pilot shortage, and the other is safety improvement. Air traffic is expected to double in the next 20 years, which would require around 700,000 new pilots. That’s a huge number! Autonomy would certainly reduce that burden. Secondly, the aviation industry has one of the best safety records in the world, and autonomous technologies could help us to further improve the safety of aircraft. With our track record of safety, Airbus is ideally positioned to drive autonomous technologies for urban air mobility with a strong emphasis on safety.

Why are urban air vehicles an excellent option for rolling out self-piloting aircraft operations?

With urban air mobility, we have a blank slate to design a safe and efficient autonomy system from the very beginning. Imagine getting into a flying taxi and being given a dedicated starting time, lane and speed that will take you from A to B in the shortest time possible with the highest level of safety and energy efficiency. That’s what we’re developing for urban air mobility.