Robot aircraft 'sense and avoid' gear in flight tests
Auto-safety kit good for droids - and flying cars
US aerospace titan Northrop Grumman has commenced flight tests of new "sense and avoid" technology which could be fitted to existing and future unmanned aircraft - or indeed to the long-awaited flying cars - allowing them to fly routinely in civilian airspace. At present, planes which don't carry qualified pilots are subject to special time-consuming bureaucracy before they can make such flights.
Flight International reports that Northrop's Detect, Sense and Avoid (DSA) suite began flying this week aboard a modified Learjet over New York State. The DSA gear includes radar, a transponder-based collision alert system, and an electro-optical camera for use by drone operators on the ground. It also has ADS-B, the new sat nav based networked airtraffic system.
"It’s not inconceivable that DSA technology will be ready for use within a matter of years," said Northrop roboplane exec Alfredo Ramirez, as quoted in Flight. Ramirez was speaking at an air safety forum in Washington this week.
Aviation authorities including the American FAA and Eurocontrol have said that unmanned aircraft will only be allowed to fly under normal procedures - so-called "file and fly", where a flight plan is filed and the aircraft takes off without further ado - if there is "the same level of safety as for manned aircraft". This is usually linked to proposed sense-and-avoid tech like Northrop's DSA, but technical requirements are sketchy at the moment.
"The real hard part is the quantification of ‘equivalent level of safety’. How does that translate into a specification?" said Ramirez.
Northrop is cooperating with the FAA on that very question in the ongoing tests, with the safety feds supplying an ADS-B equipped aeroplane for use in the flight trials of DSA.
It would appear that the FAA may want to use the unmanned-aircraft safety issue as a way to push ADS-B, a technology they are very keen on. The regulators are under massive pressure to get more flights in and out of America's overtaxed airport approaches, and a major constraint is the limited accuracy and slow update rates of current radar-based plots used for air traffic control.
US traffic controllers, however, tend to dislike ADS-B and its planned successor NextGen. These technologies offer significant potential for many more planes to be handled by fewer human controllers in future. The network might one day carry automatically-generated control instructions, as well as simple location data - leaving less information to be passed via voice traffic between controllers and aircraft pilots (or operators on the ground).
It would seem, then, that the new technology of unmanned aircraft - some examples of which are already largely autonomous in flight - may not just threaten pilots' jobs. The robot aircraft might in future menace the ground controllers' livelihoods as well.
It's not all doom and gloom, though. We and our children may lose the chance of a nice well-paid job as an air traffic controller or a military/cargo pilot (passenger airliners will probably always have human pilots, we're guessing). But on the other hand, automated and accurate high-volume traffic control - and automated piloting systems that can work within it - are one of the big things one would need for proper flying cars. Or jetpacks that were any use, come to that. Now, if only jetpacks and vertical-lift aircraft worked a bit better...
Meanwhile, it seems that if this week's calibration runs go well, Northrop's DSA rig will be tested over the summer against multiple "intruder" aircraft, to see if it can respond safely. Thus far, it has had to handle only one at a time. ®
COMMENTS
@Dave Bell
Part of making a successful system would entail not just detecting the object, but also calculating its wake effects etc ("Hmmm we have a 747 at 20000 ft doing 347 knots...) .... just like a meat version of a pilot must do.
Just missing an object should be relatively simple: take "hit it" software from a missile and just put a minus sign in front of it.
Calculate wake effects for extra credits.
The hardest part of any automated system like this is ensuring system stability and correct behaviour when the sky is full of these robot pilots. For example, how do you ensure that they don't dodge into each others paths etc.
Then, there's ensuring system stability when each vendor's software will behave differently and each different plane has different performance capabilities...
Basically a control system test nightmare, but a gravy train to keep engineers and programmers employed for the next 20+ years.
@ Dave F
> I demand that all scientists start working on jetpack technology now.
Dude! Get the with the program:
http://www.tecaeromex.com/
Tecnologia Aeroespacial Mexicana is selling one for a mere 250,000 USD (what is that now, like 30 pounds?)
They are also currently developing a bad ass personal helicopter backpack like the one from Beyond Dark Castle.
I don't know about you, but when I think personal aerospace products I think Mexican engineering.
Less ground controllers - give that to us any day
Considering how good the automated systems have become lately the less ground controllers the better. In fact we have all seen what happens when they are given the opportunity to direct aircraft without automatic assistance.
Half an hour without an automated guidance system and here we go - a plane full of kids directed into the path of a cargo liner straight away. By a human ground controller. After that everything is immediately blamed on the pilot. And if it was not for German police who got their hands on the flight recorders first the fraud would have been successful like in a large number of previous incidents here and there around "civilised" Europe.
For the reference - in most of EU the ground control and air traffic investigations report within the same business structure and if something can be blamed on the dead pilot it always is. The July 1st 2002 Bashkirian Airlines flight 2937 incident with the kids is just one example, many others. For example after the 10 January 1984 Tu 134 Berlin to Sofia the records were falsified after the crash to hide the fact that the ground crew has completely screwed up the altitude measurement for the lower bound of the clouds and blame the pilot. And so on, and so fourth.
Anonymously, for obvious reasons. And no, I am not a lizard army leutenant...
Requirements Checklist for Choosing a Cloud Backup and Recovery Service Provider
Cloud based data management
Agentless Backup is Not a Myth
SaaS data loss: The problem you didn’t know you had
What you need to know about cloud backup
