DARPA robot lands sim 73
#1
Gets Weekends Off
Thread Starter
Joined APC: Jun 2008
Posts: 3,716
DARPA robot lands sim 73
DARPA robot lands (simulated) Boeing 737
David Szondy May 17, 2017
The ALIAS system operating the 737 controls using a robotic arm (Credit: Aurora)
View gallery - 5 images A robot has successfully landed of a Boeing 737 simulator ... and it did it one handed. Built and operated by Aurora Flight Science as part of DARPA's Aircrew Labor In-Cockpit Automation System (ALIAS) program, the robot's touchdown was one of a series of flight maneuvers carried out by the system as part of the development of an automated co-pilot that can be quickly and cheaply installed in existing aircraft.
With its single, ungainly arm operating the jet simulator's controls, it may seem like an ill-conceived replacement for a human pilot. Far from it. ALIAS is a sophisticated system aimed at addressing the very real problem of the growing complexity of modern aircraft. Not only is it very difficult for pilots to qualify on an unfamiliar airplane without lengthy instruction and practice, but operating such craft can be highly distracting – especially when executive decisions are required in a hurry.
Automatic flight systems can help alleviate these problems by acting as an onboard trainer as well as a co-pilot, but current engineering procedures require either redesigning an aircraft from scratch to incorporate them, or undertaking lengthy and expensive refits that are a custom job for each mark and mod of airframe.
ALIAS is being developed to get around this. It's designed as a drop-in avionics and mechanics package that can be quickly and cheaply fitted to a wide variety of fixed and rotor aircraft, from a Cessna to a B-52. Once installed, ALIAS is able to analyze the aircraft and adapt itself to the job of second-pilot.
Along with the robotic arm, the ALIAS system incorporates an advanced tablet-based user interface, speech recognition and machine learning. Alternative versions will drop the robotic arm and provide support to the pilot by tracking the aircraft's physical, procedural, and mission state.
The idea is that, once ALIAS is fully developed it will be able to familiarize itself the aircraft within a month and take over many of the pilot's functions, allowing them to concentrate on higher level decisions and not be distracted during emergencies. In addition, it will allow for smaller crews with a subsequent drop in operating costs.
The ALIAS test was carried out in a Boeing 737-800NG simulator at the US Department of Transportation's John A. Volpe National Transportation Systems Center in Cambridge, Massachusetts. So far, ALIAS has been demonstrated in a Cessna 208 Caravan, UH-1 Iroquois, DHC-2 Beaver aircraft, and Diamond DA42 twin-engine prop plane. The latter included a demonstration of the system's ability to initiate cockpit procedures in real time as it brought the aircraft in from a simulated landing from 3,000 ft (915 m).
"Having successfully demonstrated on a variety of aircraft, ALIAS has proven its versatile automated flight capabilities," says John Wissler, Aurora's Vice President of Research and Development. "As we move towards fully automated flight from take-off to landing, we can reliably say that we have developed an automation system that enables significant reduction of crew workload."
https://youtu.be/om18cOWFL3Q
David Szondy May 17, 2017
The ALIAS system operating the 737 controls using a robotic arm (Credit: Aurora)
View gallery - 5 images A robot has successfully landed of a Boeing 737 simulator ... and it did it one handed. Built and operated by Aurora Flight Science as part of DARPA's Aircrew Labor In-Cockpit Automation System (ALIAS) program, the robot's touchdown was one of a series of flight maneuvers carried out by the system as part of the development of an automated co-pilot that can be quickly and cheaply installed in existing aircraft.
With its single, ungainly arm operating the jet simulator's controls, it may seem like an ill-conceived replacement for a human pilot. Far from it. ALIAS is a sophisticated system aimed at addressing the very real problem of the growing complexity of modern aircraft. Not only is it very difficult for pilots to qualify on an unfamiliar airplane without lengthy instruction and practice, but operating such craft can be highly distracting – especially when executive decisions are required in a hurry.
Automatic flight systems can help alleviate these problems by acting as an onboard trainer as well as a co-pilot, but current engineering procedures require either redesigning an aircraft from scratch to incorporate them, or undertaking lengthy and expensive refits that are a custom job for each mark and mod of airframe.
ALIAS is being developed to get around this. It's designed as a drop-in avionics and mechanics package that can be quickly and cheaply fitted to a wide variety of fixed and rotor aircraft, from a Cessna to a B-52. Once installed, ALIAS is able to analyze the aircraft and adapt itself to the job of second-pilot.
Along with the robotic arm, the ALIAS system incorporates an advanced tablet-based user interface, speech recognition and machine learning. Alternative versions will drop the robotic arm and provide support to the pilot by tracking the aircraft's physical, procedural, and mission state.
The idea is that, once ALIAS is fully developed it will be able to familiarize itself the aircraft within a month and take over many of the pilot's functions, allowing them to concentrate on higher level decisions and not be distracted during emergencies. In addition, it will allow for smaller crews with a subsequent drop in operating costs.
The ALIAS test was carried out in a Boeing 737-800NG simulator at the US Department of Transportation's John A. Volpe National Transportation Systems Center in Cambridge, Massachusetts. So far, ALIAS has been demonstrated in a Cessna 208 Caravan, UH-1 Iroquois, DHC-2 Beaver aircraft, and Diamond DA42 twin-engine prop plane. The latter included a demonstration of the system's ability to initiate cockpit procedures in real time as it brought the aircraft in from a simulated landing from 3,000 ft (915 m).
"Having successfully demonstrated on a variety of aircraft, ALIAS has proven its versatile automated flight capabilities," says John Wissler, Aurora's Vice President of Research and Development. "As we move towards fully automated flight from take-off to landing, we can reliably say that we have developed an automation system that enables significant reduction of crew workload."
https://youtu.be/om18cOWFL3Q
#2
Gets Weekends Off
Joined APC: Feb 2006
Position: B-737NG preferably in first class with a glass of champagne and caviar
Posts: 6,009
Robot lands a B737 Sim...
#4
Until "Control-Alt-Delete," or re-booting an Apple product is a forgotten thing, is why you don't need to worry about this: Because computers screw up all the time.
Last weekend: "Hello! Your autopilot has been encrypted! To safely land, every passenger must pay 300 Bitcoins through the entertainment system. Offer not valid if any person does not participate....use physical encouragement. If below 10,000 ft and the wifi is off, sorry."
Plus: Cat III autoland? The technology for everything except taxiing and parking has been there for years (just not wired for takeoff, but I'd guess it could have been).
Now...if the article said "Robot First Officer in 737 simulator flies hand-flown single-engine approach, raw data, manual gear!!" THEN I'd be worried.
Last weekend: "Hello! Your autopilot has been encrypted! To safely land, every passenger must pay 300 Bitcoins through the entertainment system. Offer not valid if any person does not participate....use physical encouragement. If below 10,000 ft and the wifi is off, sorry."
Plus: Cat III autoland? The technology for everything except taxiing and parking has been there for years (just not wired for takeoff, but I'd guess it could have been).
Now...if the article said "Robot First Officer in 737 simulator flies hand-flown single-engine approach, raw data, manual gear!!" THEN I'd be worried.
#5
On Reserve
Joined APC: Mar 2017
Posts: 18
Did i miss something or can it not pull back both throttles at the same time?
pretty slow too and the operating cost drop assumes there is not pilots on a ground control type monitoring station
#6
This was nothing more than an assisted auto land. I would be impressed if it "hand" flew the landing.
We can teach a real monkey (as opposed to current pilot types) to do these tasks. We can pay them with bananas and they won't threaten to unionize. All you have to do is worry about flinging poo when they get a 45 degree vector off the arrival from center.
We can teach a real monkey (as opposed to current pilot types) to do these tasks. We can pay them with bananas and they won't threaten to unionize. All you have to do is worry about flinging poo when they get a 45 degree vector off the arrival from center.
#8
Banned
Joined APC: Feb 2007
Posts: 461
Thread
Thread Starter
Forum
Replies
Last Post
SongMan
Flight Schools and Training
16
08-23-2013 06:49 PM