by E. Paul Zehr
Rhodey’s reservations about trying a solo flight in the suit are a great lead in for our next bit here. The question I want to consider going forward is, would it work? How simple would it be to just pull on the Iron Man suit of armor and do something useful with it? In that story we just talked about, the first thing Rhodey does is to accidentally demolish a wall when simply trying to move his arm. Then he mentions that it is “weird … every slight twitch becomes a great big gesture—and if I even think about moving, I do!”
The bottom line is it is really very difficult to use the Iron Man suit, especially without the direct nervous system coupling that would be required. Recall that the main underlying point for how Iron Man could ever be controlled by a human being is along the lines of brain-machine / brain-computer interfaces. This allows our thinking to “piggyback” on many concepts currently being advanced. Another summary of this overall approach, and one that also shows the potential for use in rehabilitation, is shown in figure 9.1.
Figure 9.1. Summary illustration of the goals of brain-machine interface technology. These range from more simple applications to control computer cursors all the way to linking up a human to a full-fledged Iron Man interface. The field moves forward based on advances in biomedical engineering for neural interface technology and improved understanding of plasticity in the nervous system. Courtesy Doug Weber.
Real Life! Real Science! Boy, Iron Man Would Be Really Hard to Control
Let’s take the best-case scenario and pretend that the potential user (following on from above, let’s say Jim Rhodes) is already in good mental and physical condition and has lots of experience in many different activities. You might wonder why that is relevant. Well, the better physical and mental condition means the more effective and efficient will be any interface with a machine—like Iron Man. Also, more experiences means a larger skill repertoire to draw on that can be tapped into for using the Iron Man armor. So, assuming that, what is the big deal? Well, it is one thing to just interface with the Iron Man suit. It is another thing entirely to imagine being able to control that suit while it is running, jumping, flying, shooting weapons, performing evasive maneuvers, talking to others, responding to messages, regulating temperature, and so on. Even if we just take the most basic functions that Iron Man could possibly do, like standing up and walking across a room, and then reaching out, turning a door knob, and opening a door, we are talking about months and likely even years of training.
To come up with a “ballpark estimate” of the time needed for this, I spoke again with Jon Wolpaw. Recall that we met Jon when we discussed EEGs in chapter 3. Jon’s work in brain-machine interface has largely involved brain electrical activity recorded from electrodes placed on the scalp. This interface allows for a good two-dimensional control system (think moving an image or cursor on a computer screen that has up/down and left/right as the two dimensions of movement). Such interfaces are on par with what has been achieved so far with electrodes implanted in the brain. I asked him about how long it might take a person to train and learn how to use this kind of mind-control interface. It seems that about eight to ten hours of practice time was the fastest to achieve a reasonable level of control. That training occurred in 24 sessions spread over eight weeks. Each of the three sessions each week totaled about 24 minutes of actual practice at moving the cursor in a targeted fashion. Jon advised that it is not uncommon for a person to take two or three times this time to achieve the same base level of control. They are currently studying ways to optimize the training but it is very difficult work.
Jon Wolpaw also spoke about something that I think is central to the issue of integrating a machine as intricate as the Iron Man armor with the complexities of the human brain and nervous system. And that is the issue not just of reaching some kind of level of control, or even of how fast it might take to achieve that control, but more of achieving consistent control. According to Jon, all of the methods, either invasive with electrodes right in the brain or noninvasive with electrodes on the scalp, have the same problem of “disconcerting unreliability. Control can be really good one day (or one three-minute period, or even one trial) and really bad the next.” Jon goes on to explain that the “proper conception of BCI [brain-computer interface] use is not as mind-reading but rather as skill development [that is, it’s like any other skill except it’s executed by brain signals rather than by muscles]. The problem is that this skill doesn’t become as consistent as muscle-based skills typically do. The fundamental reason for this is probably that BCI use is profoundly abnormal. The central nervous system evolved and is shaped throughout life to control muscles, period. A BCI asks it to control instead the signals (whether EEG or single neurons) from a particular cortical area. Thus, successful BCI use requires a major reorientation and a shift in the target of ongoing adaptive processes.” These points are directly in line with what we discussed earlier about how unusual it is to superimpose another “limb” or device on top of an intact nervous system (think back to our discussion about phantom limbs in chapter 4). Wolpaw’s work so far suggests that the human central nervous system “can do this (for some cortical areas), but not all that well and not very reliably.” He suggests that the best way to “address this problem may be to use signals from multiple areas (thus imitating in some fashion how normal skills are executed).” This is a really interesting idea and gets at the concept of normal neuroplasticity in skill learning.
I also spoke with Doug Weber, another leading scientist in this field. He has experience with both invasive and noninvasive systems in both human and nonhuman primates. He told me that in the monkey it is possible to train “nearby ‘patches’ of motor cortex to control arbitrary movements.” So, in this kind of experiment, monkeys could be trained to draw circles (remember this is drawing circles with brain signals, not actually using a pen or pencil!) over training with many repetitions and spanning numerous days. Doug suggests that a key issue that underlies the usefulness of machine interfaces is the quality of the brain signals that can be obtained. He said that it “is fairly easy to get good control in at least one dimension, when one or more electrodes are positioned in a brain area that exhibits strong, volitionally modulated signals—hand and face areas of motor cortex usually work very well, presumably because of the relatively large cortical map for those areas.” Doug has been working with his colleagues doing studies in patients with epilepsy who have electrodes implanted in the brain for one to four weeks.
In epileptics for whom surgery is contemplated to remove parts of the brain that have abnormal activity, doctors often implant electrodes to try to monitor and locate the source of the abnormal activity. In any case, the patients also agreed to participate in additional experiments to help with brain-machine interface. In those experiments, they “start by finding electrodes over the motor cortex that show modulated activity during motor actions like hand open/close. Once we find a good channel, it is easy to couple the output to the control of a cursor movement…. The control is not perfect, but was achieved with only a few minutes of training. Per formance improves with experience, but we have not had enough time to really study this in humans.”
Weber offers a further reality check when he goes on to say that it is difficult to really predict how complex a device—think the complexity of Iron Man—could be controlled with any such approach. That is because all the studies to date have used very simple tasks and participants have solely focused on a single task. The point is that this is “pretty artificial, since we rarely focus so much attention on simple or even more complex motor tasks. No one has really studied how BMI [brain-machine interface] performance changes when the user is performing other tasks in parallel. ‘Multi-tasking’ would require at least a portion of the BMI control to be performed subconsciously.” This last comment about multitasking is particularly relevant when we think back to our discussion of how attentionally demanding that would be in chapter 4.
The bottom line of all of this is that upon first “
jacking in” to the Iron Man suit any user would probably be able to fairly quickly (within a few hours) learn how to do something straightforward like opening or closing one hand. How long to do something even more complex like standing up and walking? Using benchmarks from neural rehabilitation and attempts at relearning skills like walking after spinal cord injury or stroke, it is reasonable to estimate that this could be achieved in about three months of training. That might give enough time to provide the ability to stand up slowly and to walk at about one-half the normal pace across a space of about 30 feet. That is pretty sobering when we think of what is shown in the comic books and movies. So, running in and grabbing the suit, throwing it on, plugging it in, is just not possible. At least, not yet.
I can see why the writers and artists show that, though, instead of a panel that says “three months later” and shows the thief slowly shuffling across the floor in the stolen Iron Man suit! Not very thrilling, but that’s the truth. Even the comic book writers did acknowledge a bit of this. Returning to the story “And Who Shall Clothe Himself in Iron?” (Iron Man #170) we began with above, Jim Rhodes does find he is having difficulty just moving around. So, he tracks down a scientist working in one of Tony’s labs for help. All we see is that some time later, the scientist tells Rhodey, “like I said it’d take months—maybe years—to dope out all that circuitry.” Umm. Yes. At the very least—for now. Just as the first “gray armor” Iron Man presaged things that are just appearing now, perhaps we will see a more intuitive control system that can be easily learned.
You might ask at this point why the suit couldn’t just remain a device that only triggers its activity from the activity of the user. Kind of like an amplified passive suit like the ReWalk we first discussed back in chapter 2. This design could work for simple and very slow tasks like walking carefully across a room with a smooth floor. However, to do sophisticated things like maneuver out of the way of Whiplash or fight Iron Monger, it would be far too slow.
Your biological body makes use of many different timescales when you move. Your neurons work with a millisecond (thousandths of a second) scale, your muscles use tens of milliseconds, your movements take hundredths of a second, and your overall impression of what is happening occurs on a “seconds” timescale. You are fully calibrated for this in the same way that your brain body maps are calibrated to your body parts as we learned in chapter 3. In order for the Iron Man suit to respond in a way that you or Tony or Rhodey could control effectively, it would need to work on command signals for your neurons on that millisecond timescale. Anything further downstream (like using actual movements of the body to trigger the motors in the suit) would introduce huge delays. Using that approach would be like trying to move around and fight Whiplash with the kinds of feedback delays you used to get on a poor transatlantic telephone connection but which is nicely mimicked now by using a low bandwidth Internet connection to have delayed Skype conversation. Everything would be out of synch. Even small delays would be catastrophic.
What Happens If Ol’ Shellhead Shorts Out?
Now let’s move on and think about what kind of real-life examples of training would be needed. So, this means we have dealt with the fairly difficult problems outlined above and assume that the suit itself can be controlled—remember this is really just saying that the user has the ability to use his or her body. Now we have to superimpose all the technical training on top of that. This issue was hinted at in the 2008 Iron Man movie in a scene where Colonel Jim Rhodes is touring some recruits through a hangar and discussing the role of pilots and advanced fighter jet technology. Rhodey says, “In my experience no unmanned vehicle will ever trump a pilot’s instinct, his insight—that ability to look into a situation beyond the obvious and discern the outcome. Or a pilot’s judgment.” Just after that, Tony Stark strolls in and says, “Colonel … why not a pilot without the plane?” Let’s pause for a minute to think about how much training just that part would take.
To think this through, let’s use qualifying as a jet fighter pilot as the example of training. Typically you need at least a bachelor’s degree. You would have to be in top physical and psychological condition. OK. Assuming you got accepted and completed officer training school, the next step would be flight school. There you would study aerodynamics, aviation physiology, engine mechanics, principles of navigation, and land and sea survival. Then, during primary flight training, you spend many hours in the air (with four solo flights) and many more in flight simulator training. That would not include additional lecture hours for flight support instruction. Upon completion of this training, the top, top candidates would go on to five more weeks of jet training “ground school,” which could include classes in meteorology, rules and safety of visual flight followed by actual flight training in aerobatics, communications, weapons use, and specialized takeoffs and landings. Next is advanced flight training, in which jet fighter pilots perform combat maneuvers and learn about flying at night. You get your wings only after five to seven years of training. That’s a lot of work. But every step here is a step that an Iron Man “trainee” would absolutely need to complete.
Iron Man is often shown having kind of standard “mano a mano” fights as well. This was clearly shown in extended sequences of hitting, blocking, punching and kicking between Iron Man and War Machine in Iron Man 2. He would need at least some basic training in martial arts—something hinted at but never really directly addressed in the Iron Man comics. It is important that he can do this, though, as his suit often powers down or malfunctions and he loses his advanced weapons capabilities. He needs at least some grounding in this.
This was cleverly shown as part of the huge Marvel-wide “Civil War” epic. In the portion unfolding mostly in Iron Man (and collected in the 2007 graphic novel Iron Man: Civil War), Tony is shown in a flashback asking Captain America for some pointers on hand-to-hand fighting. This was meant to dovetail into the early years of the Avengers and clearly highlights the status Captain America has as a dominant hand-to-hand fighter in the Marvel Universe. The images in figure 9.2 are extracted from the “Iron Man: Civil War” story and show Tony learning (panel A) and receiving (panel B) various techniques from “Cap.” Later, Tony shows that he has learned a bit from his training. I love how Cap congratulates Tony on getting it right and sweeping his legs out from under him. As Cap is falling to the floor, panel C shows him shouting, “Nice. Good follow through.” When teaching martial arts this kind of thing actually happens all the time. I always let my trainees know if they are getting it right. It is kind of bizarre, though, because really you are congratulating them on throwing you down, intercepting your attack, or applying some kind of painful joint lock!
Anyway, using a framework and timeline I outlined in Becoming Batman: The Possibility of a Superhero, Tony would need at least the initial and middle training stages. I reckon he should train in some form of martial arts emphasizing striking, kicking, and joint locking (like Cap is doing to Tony’s elbow in panel B) for five to eight years. He could do much of this training in parallel with his jet fighter qualifying. So, quite a bit of training is needed to be Iron Man. Or, check that. To be a successful Iron Man, a lot of training is needed!
All that extra training paid off handsomely for Tony in the story “The War with the Kree Is Over” found in the New Avengers: Illuminati comic miniseries that ran in 2007. In the miniseries, Iron Man, along with the rest of the Illuminati (such as Captain America, Professor Xavier, Reed Richards, Dr. Strange, Namor, and Black Bolt), is captured by the Skrull. In this particular story by Brian Michael Bendis and Brian Reed, Tony loses his armor to the Skrull and is awaiting his fate in a prison cell on board the ship. For some reason, the Skrull think it would be a neat idea to pretend to be other members of the Avengers (like Thor) to see what Tony would do. Because the Skrull are just Skrull and not super-powered Avengers, what Tony does is, well, kick butt and take names (although not so much on the names part). Once he has subdued all of the evil Skrull, he quips �
��Thanks for the combat training, Cap.” It is worth noting also that beyond the need for being able to fight in the Iron Man suit, the general body awareness that accrues with extended martial arts training would be very useful for keeping Tony Stark’s nervous system highly “tuned” and ready to go. Robert Downey Jr., who has so capably played Tony Stark in the Iron Man film adaptations, has found martial arts training to be a useful and tuning influence in his own life. He has studied wing chun, a very direct and effective Chinese martial art, for many years.
Figure 9.2. Tony Stark and Captain America training in martial arts from the graphic novel Iron Man: Civil War (2007). Note that Tony focuses on simple and effective techniques he could use while wearing his armor if his weapons systems failed. Copyright Marvel Comics.
What Is the Reality Check on Iron Man?
So, Iron Man as a possibility of a human machine certainly exists. However, the closest that we have today is more reflective of the original Iron Man introduced by Stan Lee back in 1963. The capabilities of that Iron Man are very similar to a combination of the Cyberdyne HAL and the fixed wing of Yves Rossy. Just as the Iron Man of almost 50 years ago is only now being realized, perhaps the fantastic seeming Iron Man of our day will be a reality 50 years from now.