Quadriplegic woman flies F35

[dragon]

With her mind and was only a simulator but still rather cool.

Arati Prabhakar—director of the Pentagon's advanced research arm DARPA—has revealed a breakthrough achievement in machine mind control. Jan Scheuermann, a 55-year-old quadriplegic woman with electrodes in her brain, has been able to fly an F-35 fighter jet using "nothing but her thoughts."

Scheuermann—who is quadriplegic because of an hereditary genetic disease—was recruited by DARPA for its robotics programs. Scientists and doctors implanted electrodes in the left motor cortex of her brain in 2012 to allow her to control a robotic arm, which she did successfully. But she's not using the robotic arms to control the joystick in the Lockheed Martin F-35 Lightning II simulator used for the tests. She is controlling the plane with "nothing but her thoughts," according to Prabhakar, pure neural signaling:

Instead of thinking about controlling a joystick, which is what our ace pilots do when they're driving this thing, Jan's thinking about controlling the airplane directly. For someone who's never flown—she's not a pilot in real life—she's flying that simulator directly from her neural signaling.

Quadriplegic woman flies F-35 with nothing but her thoughts

Prabhakar, who made the announcement at the first Future of War conference, celebrated last week in Washington, expressed some concern about the future applications of this technology, which was initially created to enable soldiers and people affected by motor problems, extending their bodies using robotic parts:

In doing this work, we've also opened this door. We can now see a future where we can free the brain from the limitations of the human body and I think we can all imagine amazing good things and amazing potential bad things that are on the other side of that door.

I can't see the "bad potential" of this overweighting the good potential. Sure, we can use machine mind control to remotely control killing robot soldiers on the field, but we can already do that just like we control drones.

What I can see is an incredible breakthrough that will one day enable people with motor problems to overcome whatever limitations they have. If DARPA has achieved what Prabhakar is talking about, this is an amazing breakthrough for the seamless integration of robotics and humans—and a giant leap towards the singularity.

link

[Jaepheth]

I think the bad comes when they find a way to make the control a two way street.

Very cool though.

[Pelranius]

I wonder how long it'll be until you don't need neural implants to do this sort of thing (though it's unlikely to be very precise compared to the implant method)?

[Eternal_Freedom]

I think the bad comes when they find a way to make the control a two way street.

Very cool though.

I recall an experiment done about a decade back about some cybernetics professor who had a relay implanted in the nerves in his left arm, and a similar one was added to his wife's left arm, and he was able to move her hand over the internet.

[Zaune]

I recall an experiment done about a decade back about some cybernetics professor who had a relay implanted in the nerves in his left arm, and a similar one was added to his wife's left arm, and he was able to move her hand over the internet.

I shudder to think what the sex toy industry will do with that when the technology matures a bit.

[The Duchess of Zeon]

More militarily, considering it was DARPA and a jet fighter, this is exactly the advance I've predicted will keep human pilots in the jet fighter business for the forseeable future, despite much ado about getting rid of them. Going to a direct electric control linkage instead of a two-stage electromechanic linkage will massively improve response time. Exosquad was awesome when I was a kid, and I'm going to see it happen before I'm fifty.

[The Duchess of Zeon]

More militarily, considering it was DARPA and a jet fighter, this is exactly the advance I've predicted will keep human pilots in the jet fighter business for the forseeable future, despite much ado about getting rid of them. Going to a direct electric control linkage instead of a two-stage electromechanic linkage will massively improve response time. Exosquad was awesome when I was a kid, and I'm going to see it happen before I'm fifty.

[Patroklos]

More militarily, considering it was DARPA and a jet fighter, this is exactly the advance I've predicted will keep human pilots in the jet fighter business for the forseeable future, despite much ado about getting rid of them. Going to a direct electric control linkage instead of a two-stage electromechanic linkage will massively improve response time. Exosquad was awesome when I was a kid, and I'm going to see it happen before I'm fifty.

Response time is not what is spelling the doom of manned fighters. Having to limit the performance of the aircraft to avoid killing the pilot, being able to jettison all the material requirements to host and keep a pilot alive and dedicate that weight to other things, and being able to divorce withstanding physical maneuvering force stress requirements from the recruiting/training equation greatly reducing the cost of pilots while greatly expanding the pool of those who qualify is what is doing it. That's just the remote piloting threat, we haven't gotten into automation too seriously yet.

[Starglider]

Response time is not what is spelling the doom of manned fighters. Having to limit the performance of the aircraft to avoid killing the pilot, being able to jettison all the material requirements to host and keep a pilot alive and dedicate that weight to other things, and being able to divorce withstanding physical maneuvering force stress requirements from the recruiting/training equation greatly reducing the cost of pilots while greatly expanding the pool of those who qualify is what is doing it. That's just the remote piloting threat, we haven't gotten into automation too seriously yet.

Absolutely; no current drone poses a credible threat to manned fighters regardless of control issues, and even the proposed air-to-air capable drones rely on manual control rather than artificial intelligence. On the contrary this technology can and will be applied as a more efficient means of controlling drones and will in fact be easier to deploy in that role than for manned aircraft control. However I don't think that will be enough to allow drones to obsolete manned fighters due to the perennial Achilles' heel; the vulnerability, latency and bandwidth limitations of the data link.

When we finally do get EDI-grade artificial intelligence (shut up, I liked Stealth), removing the mechanical component of human response time is not going to help (significantly). The human brain is inherently limited to >100 millisecond response times even without muscles getting involved, and computers are not, but that's not the primary issue. The main issue is that AI can (in principle) seamlessly and virtually instantly fuse data from an indefinite number of local and remote sensors into a situational awareness that is unaffected by dazzle, g-force, emotions etc. It can compare the situation to millions of stored engagements and simulate thousands of possible ways the engagement could play out in an eyeblink. I'm not talking about general AI here; we don't need anything smarter than a hawk, just scaled up, sped up and sensor fusioned into an extremely effective killing machine (that doesn't need training or leave or search-and-rescuse when it ejects or PR reaction when it is shot down).

[Sea Skimmer]

Yeah the datalink is a killer, and since any peer enemy will swat down satellites like flies, and other enemies could be fought with the B-32 bomber, that means functionally LOS UHF radio or lasers from the ground are the only option with realistic ranges of no more then about ~250nm, and about double that with an airborne repeater. That kind of distance is pretty fucking useless at this point when surface to air missiles with ranges that long existed in the 1950s. Air power is about expanding reach, anything that restricts that flexibility is a dead end since given the cost/survivability equation for modern fighters at shorter ranges expendable SSMs are bound to win out for overall effectiveness.

Super agility is also of somewhat limited appeal simply because the high G forces require very strong and thus very heavy aircraft, which then need far more power to sustain the turn, making for a very large and expensive aircraft. Whatever the rated G force a plane allows you need several times the strength to avoid an absurdly low fatigue life. Stockpiling low life drones has been talked about before, but then why not one function missiles for much less in turn? It isn't for nothing that the Super Hornet is still being produced with a 7 G limiter, and that most aircraft external stores are only rated for about 5 G. While a missile needs far more agility then the target, ERINT already basically proved that for a price a missile can have near unlimited endgame agility.

Technology like this though has far more military applications then just pilots though. For example target designation for a robotic controlled laser system. That remains the pressing issue for any automated weapon, certain target identification that is at least resistant to electronic interference. A human can get confused, but he's not going to suddenly think his entire home country is the enemy because of any single remote influence. Since we'd like robot control for lowest possible reaction time in the first place, anything we can do to shortcut the human limit is extremely useful even if it cannot match the robot outright.

Another prime place of interest is unmanned ground vehicles, and ground vehicles with robotic drivers but human commander-gunner. Using brain sensing technology in conjunction with speech recognition could be a very powerful combination and fieldable long before we can make a fully robotic tank or or weapons carrier effective. Ground combat afterall presents far greater channels then air warfare. The two stage authentication this gives to an order should greatly reduce the probability of an error in realistic situations, such as when verbal orders are being drown out by gunfire ect... it could even help correct for the human giving the wrong command vs logical intent.

Obiviously you'd also have great scope to add in other sources of information, such as eye tracking and stress monitoring, personal profiles and all the other cards one can think of. No technology should be viewed in isolation.

In any case the US is finally starting up a 6th generation fighter program office, about eight years too late, and a dedicated UCAV is NOT in the cards. But technology like this well could be.

[Marko Dash]

"think in Russian"