Aviation (in its entirety) is going autonomous, but there’s a major problem…
Protecting Humans from Themselves
When your parents used to take to the skies in the 60–70’s, there were pilots flying those planes. Today, there are still pilots flying, right?
Well, yes. But much has changed, and it’s just getting warm in here.
The introduction of jet engines in the 50’s did something interesting, it reduced mechanical failures of critical aircraft systems substantially, and our hero pilots that were once the lifeboat for our passengers, were now exposed.
Exposed to what?
A: Themselves.
Fatigue, errors, omissions, and even violations. In the Federal Aviation Administration (FAA) world, these are known as Human Factors.
Despite improving aircraft reliability accident rates began to climb:
Why?
A number of factors contributed, aside from the growth of the industry, but Human Factors role in air disasters become significantly more prominent.
Example after example of major aviation accidents can be looked at from a different perspective than they were in the 1970’s. Back then it wasn’t entirely diplomatic to lay blame on pilots. But then this happened, and 583 people perished. The world wanted answers for the worst ever air disaster at Tenerife.
It was arguably Tenerife, and this 1979 NASA report that turned the spotlight on the humans within the system and how they might operate better as a team. This gave birth to modern Crew Resource Management (CRM). Another way put, a system to protect humans from themselves.
Sadly, CRM failures continue to occur today, and will do so into the future. Air France flight 447 was one of the more prominent examples in recent times, whereby a data malfunction caused unnecessary chaos in the cockpit, such that had the pilots done nothing they would have done all they needed to do. Instead, they took the controls and flew a perfectly good Airbus A330 into the Pacific Ocean, killing all 228 people on board.
We cannot blame the pilots, they were faced with an overwhelming amount of information which their training was insufficient to diagnose. We can only blame the system. CRM had failed.
As long as the accidents continue, the system can be improved.
Paving the way for Autonomy
Since the introduction of modern CRM and the continuous improvements in aircraft and systems design aviation has been getting safer. We can largely thank the evolution of autonomy (autopilot), and the procedures forged around it. It is now uncommon for airliners to be hand flown (disengaging the autonomy) in the critical phases of flight. Good risk assessment dictates that in anything less than ideal conditions ‘hand flying’ is an unnecessary risk.
But although it may appear like we’re nearly at Zero, we’re not. The numbers still show close to 1000 people killed in 2014, a figure not exceeded since 2005, and until the human is removed from the system the risk remains.
And the problem with airspace?
Well, not much has changed since 1956 and the birth of modern day Air Traffic Control (ATC) when a Grand Canyon mid air collision occurred in un-controlled airspace, killing all 128 people on board.
This tragedy sparked the birth of the modern airspace design. Chunks of airspace tagged with a classification, given some rules and assigned a controller. Aircraft were now being actively managed.
Since then rules have evolved and classes of airspace have had iteration after iteration, but at the fundamental layer of ATC, aircraft are still controlled by humans, on the ground and in the air.
Tune in to LAX ATC at 9am on a weekday morning and what you’ll hear is something akin to ordered insanity.
Well trained, highly experienced ATC and pilots relay instructions with the other party repeating (saying a second time) the critical parts of the transaction to confirm accurate and complete understanding. The result is perhaps one instruction or airways clearance being issued every 5–10 seconds.
It’s not uncommon for airliners to speak to 5 or more seperate controllers in the final 10 minutes of flight because traffic density requires as much.
We have hit capacity.
Here come the drones
The secret is out. Drones are coming, and some big names like Amazon, Baidu, and even Dominos pizza have not been shy about signalling their intent to use drones for large scale operations.
Startups have also started steering capital to make a charge.
Drones are going to infiltrate just about every industry there is. The technological marvel that is pilot-less flight, of all sizes and performance categories brings capabilities that haven’t even been thought of yet.
What’s even more relevant, is autonomy is already migrating it’s way into commercial operations. Companies like Aerobotics, and American Robotics have already succeeded in bringing Autonomy to market. Even Airbus, who recently published their future airspace UTM Blueprint, envision;
“Drones must be allowed to self pilot … Fleets must be allowed to self manage”
Aerobotics and American Robotics are not alone in autonomy. Lockheed Martin, the US aerospace giant has already demonstrated the piloting of a retro-fitted S-76 using nothing more than an iPad.
Aurora Flight Sciences stepped it up a notch, and landed a Boeing 737.
The program is called ALIAS (Aircrew Labour In-cockpit Automation System), and was initiated by DARPA, the US Defense Advanced Research Programs Agency, and it’s amazing.
So what’s next in this innovation story?
Build the robot into the aircraft, of course.
And who might be interested in such advances in flight technology?
Look no further than aviation heavyweight Boeing, who strategically acquired the 550 man Aurora team in late 2017 to;
“accelerate development of autonomous technology”
Boeing can see it, too.
The Problem with Disruption
Firstly, traditional airline passenger numbers are set to double under conservative estimates to 2036.
And then come the drones…
The impending explosive growth introduces a number of critical problems to the aviation ecosystem:
- Inexperienced and inappropriately trained operators
- Incompatibility with existing aviation services and infrastructure
- Security risks
- Safety critical flight information distribution challenges
- Beyond Line of Site aircraft to aircraft (and aircraft to ground) separation
- Certification, Identification and Regulatory compliance risks
- Un-Trusted flight activity (public perception)
- Environmental (specifically noise) pollution, and
- Safety, to those in the air and on the ground
These challenges have lead us to the modern Universal Traffic Management (UTM) problem, a problem that needs solving not through business applications, but at the fundamental infrastructure layer of aviation.
Conclusion
Safety is of paramount importance in aviation, however the commercial reality of autonomy is weighing into the equation heavily. In order to achieve the required economies of scale to make urban air mobility and intra-city logistics a reality, autonomy is essential. These early autonomous use cases will prove the technology over time, generational acceptance will grow and full global aviation autonomy will become a reality.
Aviation is about to go through a multi decade transformation, whereby the era of manned aviation comes to a close.
“The time is now to have a global discussion about our airspace categorisation scheme. Is whether or not what we have been using for the last 50 years still sufficient?”
Doug Davis — Director of Airworthiness — CANSO (Civil Air Navigation Services Organisation)
Although many government and private entities alike are working on the UTM problem, the current thinking is narrow and provides band aid solutions at best. We believe the aviation industry is ready for a paradigm shift in thinking and in need of something bigger.
