This looks like a very promising research avenue, and I don't mean to deprecate it in any way. But this part of the article seemed a bit too pollyannish: "Currently the system hasn’t been tested on a full-scale helicopter, used instead on a rather large remote- control hobby vehicle. But as the technology fundamentally works in the same way, there’s no reason to believe this won’t make its way onto full-sized helicopters in the future."
It's not self evident to me that, at the scales necessary for use in human-piloted helicopters, the robotic legs would be light enough to avoid causing thrust-to-weight ratio problems. But as always, I'm rooting for DARPA to make it work!
Another simple point is that current rigid landing skids provide a simple surface to land with. With "legs" you know have the problem of "landing gear" which can malfunction, rigid landing skids can hardly malfunction .... oh crap 1 leg is stuck in the "out" position -- how should I land now sir? :)
> I would imagine that the ability to land on a hillside would be incredibly useful for military operations and rescue.
Perhaps, but rarely are hill sides flat, nor void of tall obstacles such as trees. Typically in hot LZ it's in your best interest to not land. By which I mean being airborne is 1 problem, having landed -- you now have 2 problems (your static and now need to take-off). This is also true typically in rescue attempts.
My understanding is that the Osprey's early issues were mechanical and training issues and not issues with the concept (propshaft issues, pilots not trained properly, poorly engineering wiring harnesses, etc) . Those things were fixed and now the Osprey is used quite a bit. I even saw one fly over my home the last time Obama visited.
Most of the current and next gen fighters have some level of V/STOL built-in as well. I think we're pretty far from what you're describing, which was maybe 15-20 years ago. Lets remember that the design team for the Osprey started in the early 80s and the original prototypes are from the 70s. The Bell XV-15 prototype first flew in '77 and then the larger version as part of this proposal flew in '83. Aircraft engineer has gone a long way since. We're able to do much more today.
I don't think this is comparable to the Osprey for the following reasons:
* the osprey's idea intent was VTOL to horizontal thrust via an adjustable wing+rotor....
--- This has NOTHING to do with feedback to external terrain
* The osprey began dev WAY before pretty much ALL the tech that would be needed to accomplish this (in scale and at low cost) even existed
* the problem the osprey was solving was NOT this....
I am note refuting your point:
"Typically in hot LZ it's in your best interest to not land." -- as this is valid...
But that does not discount the desired functionality of being able to land on non-flat surfaces in a manner which doesn't require the pilot to have gobs of flight experience and training for such a landing.
------
Let me propose a thought experiment (which I can be sure the military has already explored):
Imagine a craft that is a replacement of a helicopter in ~45 years.
It should be an airframe/device that has a set range of capabilities: it is robotic enough to land on uneven surfaces to a given angle. It can autonomously land in level or non-level surfaces... it can be sent away/recalled as a drone (drop me here - go to 25K feet and hover for an hour, then land at spot X and retrieve me - unless given further orders), it should be able to work as a swarm from fewer pilots (i.e. one pilot driving N units - either in a trailing, above, or remote vehicle/cockpit)
It should be able to have smaller advance, trailing or escorting drones tied to its safety.
it should have a body-unit (like a shipping crate) that it can either detach and leave - or come back and pickup...
the smaller escorts can swarm in groups of 6 or 4 to pickup and carry off said body-units should the main lifter be compromised....
I've seen videos of large helicopters carrying people hovering with their open tail end pressed against a hill and the front a few feet in the air, keeping the whole thing level while dropping people off on the hill, which this could have done better (less chance for error, less fuel?).
I'm surprised no other helicopters have had something like 4 linear actuators that a pilot can control to get this same effect with a bit less of the automation.
Wouldn't it be problematic to takeoff from an uneven surface? It seems like it would generate more lift from the side with higher ground and end up taking off diagonally, perhaps crashing in the process. Is there a way to compensate for that?
In hindsight, it would be hilarious (and perhaps even useful?) if the helicopter started running on its feet after landing!
Joking aside, given that the legs operate individually, I wonder what happens in case of a failure in one of them, or when one of them gets damaged during landing.
How do helicopters handle in high winds? I'd assume most rescue situations, such as rough seas and high-altitude, mountainous areas would also be very windy. I'm curious how the helicopter would hand landing on a 20 degree incline if the wind were to suddenly blow it a different direction before touchdown.
This, now obvious with hindsight, is the perfect application for other DARPA developed technologies such as the "robotic feeling hand" recently on HN. Limiting applications to anthropomorphized form factors is, of course, extremely limiting.
In addition to the use in a regular landing, I see this being a huge benefit in a hard landing. The helicopters I have been in all had seats that collapse to absorb energy in a hard landing, and these legs could do the same thing with a relatively long travel.
17 comments
[ 2.6 ms ] story [ 48.3 ms ] threadIt's not self evident to me that, at the scales necessary for use in human-piloted helicopters, the robotic legs would be light enough to avoid causing thrust-to-weight ratio problems. But as always, I'm rooting for DARPA to make it work!
I would imagine that the ability to land on a hillside would be incredibly useful for military operations and rescue.
Because the track record for "cool ways to land and take off" simply aren't that great. Eg, see the track record of the Osprey.
https://en.wikipedia.org/wiki/Bell_Boeing_V-22_Osprey#Contro...
Another simple point is that current rigid landing skids provide a simple surface to land with. With "legs" you know have the problem of "landing gear" which can malfunction, rigid landing skids can hardly malfunction .... oh crap 1 leg is stuck in the "out" position -- how should I land now sir? :)
> I would imagine that the ability to land on a hillside would be incredibly useful for military operations and rescue.
Perhaps, but rarely are hill sides flat, nor void of tall obstacles such as trees. Typically in hot LZ it's in your best interest to not land. By which I mean being airborne is 1 problem, having landed -- you now have 2 problems (your static and now need to take-off). This is also true typically in rescue attempts.
Most of the current and next gen fighters have some level of V/STOL built-in as well. I think we're pretty far from what you're describing, which was maybe 15-20 years ago. Lets remember that the design team for the Osprey started in the early 80s and the original prototypes are from the 70s. The Bell XV-15 prototype first flew in '77 and then the larger version as part of this proposal flew in '83. Aircraft engineer has gone a long way since. We're able to do much more today.
* the osprey's idea intent was VTOL to horizontal thrust via an adjustable wing+rotor....
--- This has NOTHING to do with feedback to external terrain
* The osprey began dev WAY before pretty much ALL the tech that would be needed to accomplish this (in scale and at low cost) even existed
* the problem the osprey was solving was NOT this....
I am note refuting your point:
"Typically in hot LZ it's in your best interest to not land." -- as this is valid...
But that does not discount the desired functionality of being able to land on non-flat surfaces in a manner which doesn't require the pilot to have gobs of flight experience and training for such a landing.
------
Let me propose a thought experiment (which I can be sure the military has already explored):
Imagine a craft that is a replacement of a helicopter in ~45 years.
It should be an airframe/device that has a set range of capabilities: it is robotic enough to land on uneven surfaces to a given angle. It can autonomously land in level or non-level surfaces... it can be sent away/recalled as a drone (drop me here - go to 25K feet and hover for an hour, then land at spot X and retrieve me - unless given further orders), it should be able to work as a swarm from fewer pilots (i.e. one pilot driving N units - either in a trailing, above, or remote vehicle/cockpit)
It should be able to have smaller advance, trailing or escorting drones tied to its safety.
it should have a body-unit (like a shipping crate) that it can either detach and leave - or come back and pickup...
the smaller escorts can swarm in groups of 6 or 4 to pickup and carry off said body-units should the main lifter be compromised....
/speculation
somethign like this > http://image.fourwheeler.com/f/14536367/129_0903_01_z+dana_6...
but without torsion bar and the suspension stiffening after both axle are in ground contact maintaining contact orientation
I'm surprised no other helicopters have had something like 4 linear actuators that a pilot can control to get this same effect with a bit less of the automation.
Joking aside, given that the legs operate individually, I wonder what happens in case of a failure in one of them, or when one of them gets damaged during landing.
http://www.theguardian.com/technology/2015/sep/14/robotic-ha...