The Unique Opportunity for Rapid Progress in Autonomy in Aviation vs. Automotive

As we look to the future of the aviation industry, autonomy is a clear component of continuing to safely provide efficiency at scale. This past month, Carlo Dal Mutto, Director of Engineering at Acubed for the Wayfinder team, and I had the privilege of speaking at two forward-thinking events, Re-Work Deep Learning Summit and Autonomous Vehicles Silicon Valley (AV22). Both events presented the opportunity to explore autonomy as it is applied to the automotive and aerospace industries, including their business models, rules and regulations and technology maturity. And while some people may assume all autonomous systems are equal, each use case has different challenges and requirements that set them apart.

Arne Stoschek speaking at Autonomous Vehicles Silicon Valley (AV22)

When we look at autonomy across both the aerospace and automotive industries, the biggest differentiator is that autonomy within aerospace is a natural evolutionary step with decades of progress under its belt. In contrast, it forces more of a revolution for the automotive industry. This difference creates the ongoing, forward momentum that we see in aerospace but appears to slow things down in the automotive industry.

Google started testing autonomous cars on the road in 2009, and the industry rapidly gained attention as Tesla stepped up its work in 2015. Increasing challenges around the lack of standards, intelligent infrastructure and data availability have caused headwinds to a full technology rollout. Today, when I look at the two spaces, three main categories set autonomy in aerospace and automotive apart, all of which are rooted in the fact that aircraft are fundamentally a global fleet subject to guardrails and “rules of the road.”

Safety: It’s no secret that flying in a plane is much safer than driving in a car. Safety is the aerospace industry's number one priority. Every inch of the aircraft and its systems go through rigorous testing, and strict certifications are required to meet safety standards. And each aircraft is crewed by a highly skilled and trained team responsible for the lives of all onboard.

Additionally, unlike cars, the environment that aircraft operate in - both in the air and on the ground - is more controlled, which only adds to the high level of safety. While cars are also required to meet specific safety standards, each vehicle is driven off a lot or from a garage every day by an individual who is responsible for their safety and those around them. Beyond perhaps police or speed radars, they’re not as susceptible to the same level of scrutiny minute-by-minute of their drive. And as we all know, almost anyone at least 16 years of age can drive a car.

Technology: Automated vehicle operation in aerospace has been around for decades when the very first autopilots were launched. Since then, technology and the overall demand to create automation features to improve safety and efficiency have increased. Today’s aircraft have sensors, platforms and other systems that we tap into to enhance the current setup or collect data to build new machine learning algorithms and further automate processes.

The automotive industry, which much more recently began pursuing automation, I feel underestimated the challenge at hand, which is ultimately to anticipate and mitigate human behavior, such as for drivers and pedestrians. Especially in densely populated urban environments, predicting human behavior is a complex challenge; there are simply too many possibilities that could influence the who, what, when, where, why and how of an individual's behavior. Technology can only go so far in a system where the safety of the overall vehicle network rests almost entirely on the decisions of millions of individuals or vehicles in the network.

For the commercial aviation industry, safety is achieved with a multi-layer approach, including highly standardized and regulated traffic management and deconfliction on fleet-level, well-defined airspace and vehicle-level procedures to be augmented with onboard vehicle intelligence and decision-making capabilities.

Implementation: The foundation has long been in place for autonomy in aerospace. Automated systems benefit the industry, offering the ability to improve efficiency and scalability. With more planes operating with increasing levels of autonomy, our industry will have the ability to deliver more cargo and help more people get to their destinations - as an OEM, for Airbus, this can be a driver of future growth.

A proliferation of autonomous cars will likely shift from individual car ownership to fleet utilization of cars, especially in high-density urban areas. This equates to a dramatically reduced number of people or operators buying cars. As such, the incumbent automotive industry needs not just to develop fundamentally new technology, but also to overcome several more challenges, including conceptualizing and rolling out a new business model - no small feat.

At Acubed, we have resources, infrastructure, tools and support to develop scalable software systems that are safe, efficient and reliable to build the future of flight. Our machine learning and computer vision technology, alongside our suite of sensors, can detect aircraft surroundings and support autonomous take-off and landing capabilities for aircraft ranging from small air taxis to large commercial aircraft. With data collection and analysis at our core, we are advancing the aerospace industry through this ongoing maturation towards autonomous flight.

We have seen truly progressive automation in aerospace over the past decades. That history has helped create the industry we know today, enabling an average of 45,000 daily flights transporting 4.5 billion passengers in 2019. And as we take aircraft further into the digital, automated era, the data helping “drive” those aircraft and flights provide even more valuable information that will continue to inform future product evolutions. This ongoing knowledge growth is a leading, motivating factor for the Wayfinder team.

If you’re interested in joining our team, apply here.

- Arne Stoschek