WHY THIS MATTERS IN BRIEF
Training soldiers and individuals faster gives countries an advantage in business and on the battlefield, and they’re using sci-fi like neuro-tech to realise substantial gains.
I’ve been talking about the role that new technologies, such as Brain Machine Interfaces (BMI) and Transcranial Magnetic Stimulation (TMS) devices play in helping people learn skills faster than previously possible for a number of years ago, and so far judging by the experiments, frankly, it’s the future of learning.
Take for example the volunteers who learned how to fly USAF fighter jets after these systems literally uploaded knowledge to their minds – Matrix style – which was then followed up by a $50m grant from the US military’s bleeding edge research arm DARPA who want to cut the time it takes to train spies and soldiers down from years to months and weeks. Or the Team USA Olympic ski team who managed to improve their performance by a staggering 80 percent using TMS training. And these are just two examples of many. So, it’s only natural that now the US Air Force (USAF) wants in on the action, and it believes it can use the technology to cut training times down by at least 50 percent.
This August the AFRL’s 711th Human Performance Wing launched its Individualized Neural Learning System (iNeuraLS) project, an effort to speed up pilot training using brain stimulation or as it’s also called “Neurotraining” or “Neuromodulation.” The USAF has been dogged for years by a pilot shortage and hopes the technology can help it quickly refill its ranks.
But more than quickly onboarding new pilots, the USAF is interested in the technology as a means to help pilots constantly add new skills. As the pace of change within aerial warfare continues to accelerate – for example, with the addition of Artificially Intelligent (AI) loyal wingman air vehicles like the Skyborg, and new autonomous jets that can best US top guns time and time again in dog fights, air forces that can readily learn new skills and technologies will have an edge over their adversaries.
“We’re thinking about making our workforce adaptable to change,” says Nathaniel Bridges, AFRL research biomedical engineer and neural interface team lead. “The goal is to develop and to deliver a technology that’s going to allow them to apply that [new] knowledge as quickly as possible.”
Neuromodulation is used for a wide variety of medical applications including treating chronic pain, Parkinson’s disease and traumatic brain injuries. Cochlear implants, which use electrodes to transmit signals to a deaf person’s cochlear nerve, are perhaps the best-known example of the technology.
Cochlear implants rely on the surgical implant of electrodes under the skin behind a person’s ear and are considered invasive. Elon Musk’s Neuralink, a device that the billionaire polymath wants to ultimately help people with neurodegenerative disorders get their lives back on track, as well as connect us all to AI’s in the cloud, is another example of invasive neuromodulation.
For its part, the AFRL does not believe drilling into a pilot’s skull is necessary – something that Mark Zuckerberg with his own spin on the technology also agrees with. The laboratory’s electrode earbud will instead stimulate a branch of the vagus nerve that extends to the human ear.
“It injects a small amount of current in your brain through your scalp, through your skull,” Gaurav Sharma, senior technical lead for cognitive neurocience at AFRL. “It can change brain connectivity or stimulate a particular area in the brain.”
While the earbud is pushing milliamps worth of electrical current into a subject’s brain, the AFRL will attempt to teach piloting skills using a virtual reality headset with flight controls such as a joystick, throttle and rudders. Later, the subject will be asked to demonstrate what they have learned.
“We can pull out different measures, like, ‘How accurately did you perform that manoeuvre?’” says Bridges.
The AFRL plans to compare a subject’s mastery of piloting skills to electric and magnetic field activity in their brain using a cap or a helmet that contains an array of special sensors. The laboratory will look for patterns during different stages of learning, as well as watch for changes in variables such as attention, memory and fatigue.
“Early stage learning is often characterised by more rapid performance improvements while we see more gradual performance improvements that eventually plateau in the late phase,” says Bridges. “We anticipate that iNeuraLS will help individuals move between these phases more quickly.”
A series of experiments are planned to take place over the next three years using groups of 20 to 30 non-pilot volunteers at Wright-Patterson AFB in Ohio.
“The intent is to illustrate the capability to accelerate learning in individuals who have little to no flight experience,” says Bridges.
The 711th Human Performance Wing believes the technology could also be applied for training unrelated to piloting, for example, for personnel within the Space Force, as well as workers within intelligence, surveillance and reconnaissance, maintenance and medical communities.
Using information gathered, the laboratory plans to use algorithms to correlate certain patterns of brain activity to the optimal development of piloting skills. The experiment’s machine learning algorithms will be pioneered by MIT, a federally funded research and development centre for national security technologies.
Eventually, the AFRL wants to determine the most-effective means of boosting learning, either via neuromodulation or altering the surrounding environment using immersive technologies such as virtual reality or augmented reality headsets or, as is likely, all of them.
“If they are in a state where they are most receptive to the new information that has been provided to them, then use an immersive environment to deliver the content,” says Sharma. “If they are not, then we use neuromodulation to drive the brain in the optimal state where they are more receptive to learning, and then to deliver the content.”
The AFRL believes use of realistic training tools, such as a virtual reality headset, coupled with neuromodulation, will lead to leaps in skill development.
“A combination of these technologies can lead to a multiplicative increase, and not an incremental increase, in the training rate”, says Sharma.
Overtime, the AFRL expects to find individual patterns of learning. Knowing those particularities, it believes one day it could tailor electrical stimulus and curriculum to improve the training of each individual pilot.
“A lot of technologies today are one size fits all,” says Bridges. “The reality is, we’re all different as individuals. And so, the more information we can use and acquire that relates to those individual differences, the better we can tailor the technology and learning to that individual.”
During its experiment the AFRL plans to closely control and track what is the optimal electrical frequency and amplitude for brain stimulation, and how long the effects of the neuromodulation last.
“It depends on the task that you are doing,” says Sharma. “In some of the studies, we found that when we did stimulation for two hours the effect lasted for like 16 days.” In another clinical setting that aimed at treating a bleeding disorder the effect lasted six months, he adds.
The scientists say past experience with neuromodulation experiments and strict procedures will keep subjects safe from harm.
“In case we see something bad happening, which I don’t think we will, we will have the ability to shut it off because it is controlled in a closed-loop fashion,” says Sharma, noting the real-time brain activity measurements. “We have implemented a number of these technologies in our past projects on similar ideas of improving performance, so their safety is very well established.”
The iNeuraLS experiment is to culminate in 2023 with a demonstration where participants will train to perform different flight manoeuvres with a virtual reality-enabled flight simulator. The ultimate aim is to show a reduction in the amount of time it takes for the participants to learn the skills compared to a control group, says Bridges. The AFRL team also plans to create an integrated brain stimulus and mapping device by the end of the project.
“Ultimately, we will be able to translate this technology into something that can be used in an operationally relevant environment, like in an aircraft,” says Sharma. “However, there are different things that we need to consider, for example, are these materials going to hold in the high-g environment that an aircraft has to go through.”
Current systems for mapping the brain’s magnetic fields, called magnetoencephalography systems, are large throne-like devices with massive, bulbous helmets that descend over a patient’s head and restrict movement. The equipment also has to be chilled and stored inside special rooms that stop the earth’s magnetic field from interfering with measurements of the brain’s magnetic fields.
To shrink the technology, the 711th Human Performance Wing is working with the AFRL’s Materials and Manufacturing Directorate and start-up Sonera Magnetics to develop a new compact sensor array for mapping the brain’s magnetic fields.
“They are coming up with new materials that will allow us to get the immediate signal with good spatial and temporal resolution, but still in a wearable, portable format, which may not require that kind of a shielded room to record that activity,” says Sharma.
The 711th Human Performance Wing is also partnering with Microsoft, which will manage content delivery and will provide guidance on changes in virtual and augmented reality platforms and hardware. Teledyne Technologies will integrate the brain mapping and stimulating components into one device.
Ultimately, AFRL scientists hope the iNeuraLS project is a stepping stone toward a mind meld between man and machine, something Elon Musk calls AI Symbiosis – a futuristic application that might enable a pilot to control loyal wingmen using nothing more than his brain, and an application that could be the on ramp to turning pilots into organic computers, which is as sci-fi as it sounds, where they themselves become a hybrid human-machine computing node.
“If we can validate some of the core components of the technology that we are pursuing in the project, that can have applications beyond training,” says Sharma. “Can this improve the communication or the symbiotic [relationship] between the man and machine?”
Yes, probably… actually, just yes. Yes it can. And who can’t wait for that awesome future!? #Crazy