Tomorrow Soldier: How The Military Is Altering the Limits of Human Performance
Imagine a group of volunteers, their chests rigged with biophysical sensors, preparing for a mission in a military office building outfitted with cameras and microphones to capture everything they do. “We want to set up a living laboratory where we can actually pervasively sense people, continuously, for a long period of time. The goal is to do our best to quantify the person, the environment, and how the person is behaving in the environment,” Justin Brooks, a scientist at the Army Research Lab, or ARL, told me last year.
ARL was launching the Human Variability Project, essentially a military version of the reality-TV show Big Brother without the drama. The Project seeks to turn a wide variety of human biophysical signals into machine-readable data by outfitting humans and their environment with interactive sensors.
The Army is not alone. The Air Force, Marine Corps, Navy, and their special operations forces are also funding research to collect biophysical data from soldiers, sailors, Marines, and pilots. The goal is to improve troops’ performance by understanding what’s happening inside their bodies, down to how their experiences affect them on a genetic level. It’s not exactly genetically engineering soldiers into superhero Captain Americas; the U.S. military insists they have no intention of using biometric data science for anything like the genetic engineering of superior traits. But it’s close. The military is after the next best thing.
If today’s Pentagon leaders get their way, the next generation of fighter jets, body armor, computer systems, and weapons will understand more about the pilots, soldiers, and analysts using them than those operators understand about the machines they are using. The very experience of flying the plane, analyzing satellite images, even firing a gun could change depending on what the weapon, vehicle, or software detects about the person to whom the weapon is bound. To make this dream real, Pentagon-backed researchers are designing an entirely new generation of wearable health monitors that make Silicon Valley’s best consumer fitness gear look quaint. They’re discovering how to detect incredibly slight changes in focus, alertness, health, and stress — and to convey those signals to machines. Design the boots well enough and the super soldier will arrive to fill them.
Army Research Lab researchers already monitor individual subjects from six months to two years. Brooks wants to expand that to other military training environments, such as the U.S. Military Academy at West Point, and then to more than a dozen universities. He hopes the data will reveal how people of varied size, weight, height, health, level of alertness, etc., differ in terms of the signals they send out — hence the name “human variability.” That, in turn, will help researchers gather much more precise information on how different people interact with their environment. The ultimate goal is sensors that can tell the Pentagon how each human soldier performs, or could perform, to their best ability, from battlefield to homefront.
“It's not just while they're at work, but also when they go on leave,” says Brooks. “This is continuous, with the highest practical resolution that we can obtain for a long period of time. Hopefully, we would see information going into many programs” to build future gear. “A greater understanding of natural human variability would then feed pretty much any system that adapts to the person.”
It’s an ambitious undertaking, considering the current limitations of body-worn sensors. Over the past two years, the military bought more than $2 million worth of FitBits and other biomedical tracking devices. But it turns out that off-the-shelf consumer devices aren’t good enough for the military’s biotracking ambitions. So researchers are creating a new class of wearables, based on new research into embedding electronic components into fabric. If the electrodes are too small, the signal is worthless; too big, and they feel like an artificial electric shell separating the wearer from the real world. The connection between the environment and the human must remain seamless.
One application for such sensors is helmets that record brain activity while their wearers do their jobs. An ARL team is preparing for continous electroencephalography, or EEG, by using 3-D printing to create helmets that fit perfectly to each individual soldier’s head. But the military is not eager to embed wires and metal into gear that’s meant to protect a soldier during a massive blast. So the lab is constantly looking at new materials, solutions, and tradeoffs, inching toward sensors that collect information without getting in the way of soldiering. Lab technicians showed me one experimental electrode that they were making that was so small and soft to the touch it seemed to have no metal in it at all (they are in fact constructed of nanofibers that conduct electricity, encased in silicon.)
Merged and Monitored: The Fighter Pilot of the Future
The F-35 Joint Strike Fighter Helmet from Rockwell Collins. Photo courtesy of Rockwell Collins.
The Air Force, as well, needs a next generation of wearables to help tomorrow’s combat aircraft understand their pilots. Modern fighter jets expose human bodies to physical forces that are still not entirely understood. In 2010, multiple F-22 pilots reported in-flight episodes of confusion, shortness of breath, and skin-color changes — all symptoms of hypoxia, or decreased oxygen in the blood. The reason was speed.
“I pull a G in the airplane, blood has a tendency to collect in some of those dependent areas of the body, like the arms and legs and that,” said Dr. Lloyd Tripp, a program manager for aerospace physiology and toxicology at the Air Force Research Lab’s 711th Human Performance Wing. Two years later, the Air Force began to affix sensors inside the helmets of F-22 pilots to read the blood-oxygen level of their temporal artery.
Around the same time, the Russian military was also seeing confusion and skin-color changes among their pilots who pulled high G-forces, Tripp said. Lacking the same sensor technology, Russian commanders began to give pilots blood transfusions before their flights. It didn’t work. Russian pilots flying at supersonic speeds suffered hypoxia at greater rates. “They didn't actually admit that for quite a few years,” he said. Correct diagnoses enabled the U.S. Air Force to read the problem and improve performance.
Beyond helmets, Air Force researchers are working on what they call a comprehensive cognitive monitoring system. This means exploring what sensor technologies work well for what purposes, and what signals can be detected without interfering with or disturbing the pilot — who is, after all, supposed to be flying a combat mission. Depending on what you seek to measure, they found, you may no longer need a physical sensor on the body. You can now collect incredibly intimate and important internal health data with cameras.
Take cerebral oxygenation, the amount of oxygen in the tissue of specific portions of a pilot’s brain. You can measure this key biophysical signal by shining infrared light on the forehead because the blood in front of the skull is about as oxygenated as the brain tissue behind the skull wall. “If I'm shining that infrared light through the skin, I can see the amount of oxygen within the blood in that tissue. As I increase G-force, I'm decreasing the amount of oxygen that I have here and that decrease in oxygen is directly correlated back to decreases in cognitive function,” said James Christensen, a portfolio manager with the 711th Human Performance Wing.
Another research project configured simple laptop-camera lenses to detect whether a person’s hemoglobin is oxygenated, which makes blood shows up slightly redder, or de-oxygenated, which is slightly bluer. Essentially, this lets you read a person’s heart rate from a distance.
Even your breath says something about your physical state. “The ratio between oxygen and carbon dioxide will change as I become more and more fatigued. That's important because as I'm fatigued, it takes about 24 hours for me to actually recover 100 percent,” Christensen said. “That fatigue is important because my muscles can't strain to push the blood back to my head and so the probability of me losing consciousness increases significantly.”
Good sensors can even detect changes in metabolism that indicate weariness and stress before the person notices. When you’re stressed, you exhale fat — or rather, water-soluble molecules called ketones that your liver produces from fat. Stress is detectable by the molecular content of your breath.
“We’re working with some folks over at our materials lab and they have a couple of companies that are looking at sensors that are going to be placed in the [pilot’s oxygen] mask that'll look at those types of fatigue-related volatile organic compounds,” says Christensen.
Your eyes, too, give you away. “Imagine eye-tracking cameras,” Christensen said. “If those can collect not just the motion data and the eye-motion data, but those are also getting heart rate and respiration, then we can have no hardware on you at all and still get all the same physiological metrics…A certain amount of cognitive workload tends to correlate pretty highly with stress generically. You can combine heart rate with several other measures to get at workload stress; vigilance, even."
"We are comparing it, just for reference, with wet medical electrodes on the chest. Under most conditions, you can do about as well as wet electrodes,” he said. The lab is “testing the limits of how far away can you get and still get a reliable signal. It turns out, it's mostly an optics problem.” That means cameras and lenses alone can detect those subtle changes in stress and attention. It’s just a matter of figuring out which ones.
There are privacy ramifications to collecting so much information. A simple camera can gather enough biometric data on an individual to understand how small changes in heart rate can be a sign of stress. For a fighter pilot, an analyst, or a soldier, this might help warn of decreased cognitive ability. But among the general population, stress can also be a signal of deception, depending on the context in which that stress expresses itself, such as an interview at a checkpoint. Today’s military-funded biophysical research shows that it’s possible to detect that stress response from 100 meters away, and perhaps even at longer distances. In theory, if you could create a lens that could capture infrared data at sufficient resolution (currently, only a theoretical possibility), you could measure brain tissue oxygenation from low-earth orbit. You could see stress from space.
When performed without a subject’s awareness or permission, biophysical monitoring can be a violation of privacy. But conducted as part of an experiment with knowing volunteers, like elite soldiers eager to understand their bodies and improve their own performance, it becomes a powerful tool. One former special operations training psychologist, who currently works for a major league baseball team, said the elite soldiers he had served with were eager to improve their performance through data. In the Air Force, pilots want to improve how they fly, complete their missions, interact with their equipment, etc.
Bit by bit, this science is making its way into actual gear and weapons. In the year 2020, Navy SEAL teams and Army Rangers could take down high-value targets while wearing an exoskeleton that’s earned the nickname ‘Iron Man.’ Biophysical sensors will play a big role in the way the suit functions.
In February and March, the Air Force successfully tested a new helmet with “physiological monitoring capabilities,” as Tripp put it. Its heads-up display shows different information based on how the pilot is feeling and other factors. The goal is to give every pilot a slightly different experience based on their unique physical and mental strengths and weaknesses, as well as their physical condition at the moment. Lab researchers and contractors anticipate it will guide the design of the next U.S. fighter jet, to be launched between 2025 and 2030.
“I may do a really, really good job on a spatial cognitive task where I'm looking at a radar warning display, and maybe James doesn't,” Tripp said. “The thought, down the road, is to quantify my performance in these decreased physiological conditions from a cognitive perspective, and then use the changes in physiology to make the airplane smart about what kind of help I need.”
Kaleb McDowell, lead of ARL’s Center for Adaptive Soldier Technologies, said there will be a fundamental give-and-take when designing the weapons of the future. People perform better when their tools are crafted specifically for them. But it’s hard to design for individuals quickly and at the scale of hundreds of thousands of troops. That’s why the design of weapons software today flows toward averages – and mediocrity. “You’re designing it to be simple for everyone,” McDowell said. “A guy that's great spatially doesn't use the spatial capabilities on any system that you see today. A woman that has a great math capability isn't using that in today's systems because no one's conceiving of a system that actually relies on that capability. You just design it for everyone to use.”
So McDowell wants to build weapons that adapt to their users. “I want my system to be able to rely on, say a great memory, poor math capability, and a great spatial capability. I want the system to be able to say, ‘This person's really creative. How do I tap into that imagination when doing this dull task?’”
But that also affords the military far greater insight into what job or mission they are giving to what soldier. Researchers say that that is a key benefit of the new data-collection programs. “The basic goal here is: we want to get greater precision and accuracy in predicting which people will succeed in particular job areas or missions,” Air Force research psychologist Glenn Gunzelmann said at a National Defense Industrial Association event in March.
You Can Be Programmed and Unprogrammed
What if the Air Force could use an airman’s personal history to predict how he would perform in his surroundings – even in battle? The military already keeps massive records on troops’ lives that, if structured properly, might furnish a treasure trove of mineable health data.
Col. Kirk Phillips, associate chief for bioenvironmental engineering at the U.S. Air Force, and his colleague Dr. Richard Hartman are pioneering a program called Total Exposure Health. The goal is simple: collect and analyze as much data as possible about what happens to soldiers beyond the battlefield, right down to the kinds of molecules to which they are exposed. And in the military, a lot is recorded.
“In the Air Force, for instance, if you want your house treated for an infestation, that gets recorded,” Hartman said. “We have more opportunity to interact with them [people in the military] in that total environment. Where they live and where they work. It’s something that is better known to us. They receive health care from us. We can measure their exposure at work so we can offer to measure their exposure at home. We can know what exposures are in the environment because nobody is saying, ‘Why are you measuring the amount of chemicals in the soil?’"
If you could take that information and convert it into structured data, algorithms could produce all sorts of new insights about how individuals are interacting with their environment, in real time and in incredible detail. Phillips believes that exposure science has enormous applications in the emerging field of epigenetics research.
Here’s where Phillips’ vision becomes both revolutionary and a controversial.
Epigenetics is what your genes do with the change that you experience. It’s based not on your immutable DNA, but rather on your micro-RNA, the tiny molecules that turn on or off in response to stimuli. Think of a stress hormone that your body creates in response to an event. When your stress level goes down, new micro-RNA are formed and that controls gene expression in everything from your metabolism to how well you recover from disease. But it’s incredibly difficult to understand these interactions, precisely because everyone’s genetic makeup is different. Phillips hopes Total Exposure Health will yield a fuller picture of how specific sets of experiences affect specific sets of micro-RNA inside a specific soldier.
"Let's say that external stress happens to be a chemical exposure you may never encounter, or there may not even be a micro-RNA that turns that part of your gene off that it activated. You may have the gene that activates under that exposure, and I may not. You may be very susceptible to a chemical that I have very little susceptibility to,” Phillips said.
Phillips thinks that if he can detect these kinds of things for the military, Total Exposure Health could revolutionize civilian healthcare as well. It offers high specificity on individual health on a scale of billions of people.
“You've probably read in the newspaper that they did a big study and they looked at red wine. They tried to see whether there was a health benefit to drinking red wine. Another study says: Maybe. Another study says: Not really. That's because it's population-health-based,” he said. “They're just trying to pick a random population to see a population level change. If you have a gene that's not very prevalent in a population, then you won't get a population result of that exposure. Precision health and medicine says, ‘I should understand your gene in a way that I can understand whether your gene is activated by red wine and whether that activation is a health benefit, or health detractor.’”
“Right now, you might go, ‘I've read studies about red wine and they seem to be all over the place. I don't know whether I should do it.’ Health care of the future would look like this: your physician would say, ‘You know what? We looked at your genome. We know that red wine activates in the genome in a way that provides the health benefit. You don't have the gene, so only drink red wine to the level that you find it pleasurable in social situations.'"
If Phillips is right, Total Exposure to Health may ultimately give millions of people an incredibly detailed understanding of how their health choices affect their future. Not just, for example, how much alcohol is unhealthy for an average person of their age, weight, etc. to consume, but how much red meat, caffeine sleep, etc. is good for them specifically.
It is becoming possible to know the health outcome of any action with an accuracy that would have seemed supernatural just a few years ago. The ability to comprehend the probabilities that form the future is the ability to influence it. The interplay of our genes and our experiences, of nature and nurture, moves from the mysterious to the knowable, or at least toward the more knowable.
Designing for the Soldier of Today and Tomorrow
For the military, this opens up new choices that are pulled directly from dystopian science fiction: anticipating what soldier is best suited for what assignment or mission.
In 150 BC, the Greek writer Polybius observed that Roman military units were doing something that no known army had done before: keeping careful and consistent records. The Romans could ration grain and wine across soldier classes and types because they had a uniform system of recordkeeping for just that purpose. The reduction of unpredictability was proving a great battlefield advantage.
Imagine a military doing the same thing today but on a level both grander and more granular, where the substance to be rationed is a particular type of soldier personality, or even a specific kind of neurotransmitter.
Dr. Josh Hagen of the U.S. Air Force's 711th Human Performance Wing briefs then-Defense Secretary Ash Carter on wearable sensor technology during a late 2016 visit at Wright-Patterson Air Force Base, Ohio. (U.S. Air Force / Wesley Farnsworth)
Again: U.S. military officials are adamant that they are not genetic engineering military personnel and have no plans to do so. But they do not expect potential adversaries to share the same constraint, especially if it offers advantage over the military might of the United States. (Remember the movie Rocky IV? Just consider the Russian government’s recent systemic and secretive use of performance-enhancing drugs to win at the last Winter Olympics. Now imagine a battlefield of soldiers.) It’s a future to either embrace or learn to defend against.
If you were to use biometrics to genetically design a superior military, how would you do it? The outlines are visible today.
Individuals disposed toward risk-taking are probably better suited for particularly dangerous deployments and missions. But those same individuals are poorly suited for other aspects of military life, or less exciting military vocations, according to a landmark 2000 study by U.S. Army Maj. Michael Russell. He proposed that there were two primary military personalities: soldiers who exhibited a need for action and unpredictability (high stimulus-seeking) and people who were attracted to the military because its life offers a high degree of structure and discipline. A military needs both types to perform at peak but therein lies a fundamental contradiction. Military life is incredibly structured. War is unstructured. The stronger your attraction to one set of stimuli on the spectrum, the greater your aversion to the other.
“It has long been recognized that a peacetime army differs in many ways from that of an army at war. This is intuitively obvious: destruction of personnel and equipment, even enemy equipment and personnel, is somewhat antisocial,” Russell wrote, achieving a new plateau in euphemism by calling blowing up the enemy "somewhat antisocial."
“To plan the ultimate defeat of an entire army or nation on the battlefield requires at least a dose of narcissism. Therefore, those personality attributes that make for a war hero are primarily from cluster B. These people do not function as well in garrison. Such individuals thrive on challenge and require constant stimulation,” he wrote.
Merle Parmak, a military psychologist and a former Estonian Army captain, discovered that individuals who perform better in a highly structured, less exciting environment can also have great military careers, but perhaps not on the front lines. To a certain extent, you can train risk-taking soldiers to better accept the rigid boredom of military life away from the action, just as training can help structure-minded military personnel to better cope with the unpredictability of combat. But sticking the wrong person in the wrong job has costs.
Now consider the role that dopamine plays in risk-taking, according to an established and rapidly growing body of research. Dopamine levels are at least partially controlled by the monoamine oxidase A gene, or MAOA. A specific variant of MAOA called VNTR 2 was correlated with violent antisocial behavior, but only in the context of a stressful life event in adolescence.
If the connection between genetic factors, life experience, and risk-taking can be better observed, can they also be controlled? This is the question that will loom over military leaders in the decades ahead.
The Pentagon’s projections for future conflict are these: highly confusing and stressful urban warfare engagements. Population demographics pushing people into megacities means more door-to-door fighting, and more rules to protect civilians against adversaries who don’t have the same commitments to internal law or norms. War in the future … sucks.
Depending on the intensity level of different conflicts in which the United States is engaged, the level of violence, the effectiveness or the simple ruthlessness of the enemy, the military may feel pressure to keep up with an adversary short on the reservation. Should the United States find itself in such a conflict, Pentagon leadership may feel very differently about genetic engineering to secure better soldier performance, especially doing so might degrade U.S. military advantage at less cost.
Should some future leader – of any country – make the decision to abandon the ethical frameworks we live by today, the tools will be there for him or her to make a swift transition.
But even genetically engineered humans might lose the battle in the end. The pace of war exceeds the speed at which humans can observe what’s happening, conceptualize a strategy, and deliver commands to pull off complicated counter-maneuvers. This is sometimes called the observe, orient, decide, and act, or OODA loop, and it’s moving from a thing that humans do on the battlefield to a thing machines do. If you listen to the Pentagon’s top strategists when they talk about the future, this concern rises repeatedly. “When you think about the day-trading world of stock markets, where it’s really machines that are doing it, what happens when that goes to warfare?” William Roper, the head of the Pentagon’s Strategic Capabilities Office, asked at last year’s Defense One Tech Summit. “It’s a whole level of conflict that hasn’t existed. It’s one that’s scary to think about what other countries might do that don’t have the same level of scruples as the U.S.”
Given a choice between losing a major conflict and taking advantage of next-generation science to create a new advantage, it’s not hard to predict what any military will choose.