The Joint Fires and Effects Training System, created by the University of Southern California’s Institute for Creative Technologies (ICT), has received some attention for using what its makers describe as ‘the creative talents of Hollywood and the game industry’ plus ‘movie-industry effects’ to mix simulation with ‘story-driven content’.
The ICT was established in 1999 as a defence-affiliated research lab following a report published in 1997 as ‘Modeling and Simulation: Linking Entertainment and Defense’.
ICT has since also produced UrbanSim, a game that trains in ‘complex counterinsurgency and stability operations’, and Full Spectrum Warrior, an entertainment game that grants users tactical command of a light-infantry platoon.
The JFETS trainer is one of many dismounted forward-observer simulators that allow users to practice calling for indirect fire (artillery and close air support). Similar generic products are freely available online for entertainment purposes.
The more broad trainers, ‘synthetic battlespace rehearsal systems’, are themselves modified from off-the-shelf commercial games. Virtual Battlespace Systems (VBS) 1 and 2 were based on a commercial game called Operation Flashpoint: Cold War Crisis, adapted by Bohemia Interactive Australia and distributed worldwide by the Florida defence contractor Coalescent Technologies.
Meanwhile all four branches of the US armed forces, together with the DoD, Homeland Security and other government agencies, have pooled training-systems resources and commands in a ‘unique military collaborative alliance supplemented, supported and augmented by academic and industry leaders in the modeling and simulation, human performance, and training domains.’
This sixty-year-old alliance between defence contractors, universities and military is now called Team Orlando, and is based in the Central Florida Research Park.
The latter is home to the National Center for Simulation. It also contains the headquarters of IDEAS, a subsidiary of entertainment giant Disney, an ‘innovation studio’ that produced products for the US Navy’s ‘Revolution in Training’. Lockheed Martin (which produces a series of games under the DARWARS programme) has its regional office there, as do Boeing, Northrop Grumman, BAE Systems (the world’s largest military contractor) and videogames companies EA and n-Space.
Another resident firm, AVT Simulations, lists as a key activity its adaptation of commercial game engines to meet the training requirements of military customers.
Each year Team Orlando holds and participates in events like I/ITSEC (the Interservice/Industry Training, Simulation and Education Conference), GameTech and Otronicon.
These are intended, firstly, to facilitate cooperation between armed services, government agencies, academia and private firms in developing military training products and services. They are also supposed to showcase ‘how some artists are crossing over from military to entertainment to military applications’, and to attract ‘current and future gamers to add the defense industry as a career option.’
At last year’s GameTech Will Wright, creator of the Sims series of games, gave the keynote speech; this year the honorary address was delivered by the head of game-development firm Zombie Studios. The 2011 event was held during the same week in late March at the same Orlando hotel as the Army Games for Training Conference.
Indeed, insofar as the US Defense Department and the various armed-forces agencies take a serious interest in games and virtual simulators, it almost exclusively is as training instruments that can meet much the same needs as do live exercises.
Altogether, the US Army’s Warfighter Training Alliance, led by military contractor Raytheon, integrates three kinds of training: Live, Virtual (‘immersive’ simulations, operated by Computer Sciences Corporation) and Constructive (game-driven simulations, administered by General Dynamics).
Wright, in his keynote address to GameTech 2010, listed a bunch of trendy topics (network theory, complex adaptive systems, cellular automata, power-law distributions, chaos theory, nested hierarchies, etc.) in which, he claimed, the concerns of games design and those of counter-insurgency planning intersected.
But Wright’s talk was followed immediately by some quick words from Frank C. DiGiovanni, Director of Training Readiness and Strategy at the Pentagon’s Office of Personnel and Readiness. DiGiovanni is the policy coordinator and overseer of training for all four branches of the US armed forces; his office was the co-sponsor, together with Team Orlando, of GameTech.
DiGiovanni oversees development of Advanced Distributed Learning (ADL) initiatives.
ADL emerged during the Clinton administration as an attempt to ‘support geographically unconstrained training’ in a future operating environment of ‘persistent conflict.’ The ADL website asks: ‘Why is the DoD putting so much effort into advanced learning technology?’
The increased rate of deployments of the DoD’s forces in recent years, which often involve rapid, unplanned movements to locations around the world, highlights the need to provide training on demand to individuals and units deployed worldwide. Accordingly, because of more demanding deployment criteria and other time-sensitive constraints, the DoD recognizes that yesterday’s framework of “right time, right place” learning, with its use of set times and places, may not meet future requirements. It also recognizes that providing “anytime, anywhere” instruction is essential to maintaining military readiness in the information age; future forces and their support activities must be highly adaptive to address threats effectively and rapidly.
The Pentagon’s 2010 Strategic Plan for the Next Generation of Training declared that one of the two training revolutions to have occurred in the previous 40 years was ‘the development of large scale ranges that allowed large force exercises and high fidelity conflict simulation in the late 1970s, early 80s.’ This referred to the giant National Training Centre at Fort Irwin in the Mojave Desert.
But this site was no longer adequate. Parts of it had recently been leased out for solar thermal and PV installations, and as a fixed location it no longer met strategic needs, which tended towards military expansionism and global power projection. This would leave hundreds of thousands of US soldiers and marines deployed in combat zones and stationed at overseas garrisons, rarely rotated back home:
Not only are U.S. forces deploying more often, they are experiencing an increasingly diverse range of missions when they do…As current operations, major materiel acquisitions and personnel related expenses continue to consume an increasing portion of defense budgets, funding for training will face significant budget pressures. Training facilities and other resources may often be underfunded, especially in relation to the increased demand described above. Meanwhile, competing demands for air, ground and maritime maneuver space, the growing reach of sensors and weapons, and environmental constraints on the use of existing ranges and maneuver areas increasingly challenge our ability to conduct live training, especially on a large scale….The long-term objective is to produce an immersive training environment… This environment must be…globally accessible 24/7…Because DoD is a global enterprise, access to that environment should be persistently available via the Global Information Grid (GIG) to the full range of users.
The 1999 DoD Strategic Plan for Advanced Distributed Learning put it thus:
In tomorrow’s dynamic threat environment, America’s forces may have to deploy on a moment’s notice…They must continuously learn, simulate, and rehearse, whether they are in school, at home station, at home, en route to, or in the theater of operations…While yesterday’s right-time, right-place learning paradigm met yesterday’s military requirements, it can not meet future requirements based on more demanding deployment criteria and other time-sensitive constraints…[Meeting] the more complex readiness needs of the future will require the Department of Defense to re-engineer its ability to deliver learning to an anytime-anywhere objective.
DiGiovanni thus made clear that the ‘strategic trends and pressures’ turning military instructors towards games and simulation were ‘high demand on our live-training resources’ from ‘competing demands.’ In such circumstances, games and simulations were a direct subsitute.
So just to be clear: what ultimately has prompted the Pentagon’s turn towards games and simulations as ‘distributed’ trainers? What lies behind the growth of institutional, funding and personnel links between armed services, giant defence contractors and private entertainment firms in the area of modelling-and-simulation R&D? (There are, of course, countless other examples of broader integration between entertainment firms and the DoD, from recruiting tools like the Army Experience Center and America’s Army, to the Catalyst workshop for aspiring screenwriters.)
According to the policymakers themselves, it is less the demands of network-centric warfare than the state of US grand strategy.
The latter was described two weeks ago by the Harvard International Relations theorist Stephen M. Walt on the Foreign Policy blog: ‘Most of the U.S. foreign policy establishment has become addicted to empire, it seems, and it doesn’t really matter which party happens to be occupying Pennsylvania Avenue.’ As much has been clear at least since the Clinton Administration, the first to openly proclaim the territorial sovereignty of other states to be a revocable licence, granted by the “international community” (for which read Washington and whatever alliance it saw fit to cobble together) and enjoyed only at the latter’s pleasure.
The active citizens of the US polity – the business class, the political class itself, the Security State, the echo-chamber of thinktanks and journalists – are committed to the establishment and retention of military protectorates over the world’s energy resources, sealanes and other zones of strategic importance as a means to prevent the emergence of peer competitors, maintain the position of US firms, preserve the dollar’s privileged position, etc.
In such conditions, the dispersed masses of US ‘Warfighters’ – hundreds of thousands of them penned in fortified garrisons or engaged in combat operations around the world – can no longer be assembled reliably and periodically at home bases or live training ranges. They must find new tools by which to learn and retain their skills.
The US Army’s Research Institute for Behavioral and Social Sciences (ARI) undertakes research into training and personnel matters; its mission is to ‘maximize individual and unit performance and readiness to meet the full range of Army operations.’ Much of its recent work explores the worth as combat training exercises of games, virtual ‘immersion’ and simulation.
The psychology department at the University of Central Florida (UCF, part of the research cluster together with Team Florida) has also devoted itself to such questions. Psychomotor combat performance (marksmanship, shoot/no shoot decisions, etc.) is known to diminish with the application of stress (noise, sleep deprivation, enemy attack). Increased heart rate, sweating, shallow breathing and higher cortisol levels may lead to ‘unpleasant or intrusive physical or emotional reactions’: tunnel vision, auditory blocking, distraction from task-relevant stimuli, muscle tension and rigidity, and ultimately freezing and non-execution of the task itself.
According to research published in the journal Military Psychology in 1989, based on studies of Israeli veterans of the 1982 Lebanon invasion, the strongest predictor of combat stress reactions (which manifested clinically as compulsive crying, sleep disturbances, screaming, vomiting and diarrhea, numbing or levelling of affect, trembling, paralysis, withdrawal from surroundings etc.) was the sight of death (whether that of a comrade, enemy or officer).
Given that inflicting enemy casualties is the basic measure of combat proficiency, this presents a problem.
According to research from UCF psychology professor Eduardo Salas and James E. Driskell from the Florida Maxima Corporation, ‘the primary purpose of stress training is to prepare the individual to maintain effective performance in a high stress operational environment.’ The first way of inoculating against stress reactions, and thus making combat skills resistant to stress-induced degradation, is to ‘enhance familiarity with the target environment.’ This will render an ‘aversive event…less novel and unfamiliar.’
Familiarity is provided by preparing the trainee with the relevant sensory information in conditions that realistically approximate, to a reasonable degree of fidelity, the task/stress environment. The second way is by ‘overlearning’ the relevant skill or task, so as to make it a routinized habitual response to the context or stimulus.
Various DoD training outlines have emphasized that simulations, games and immersive trainers must produce ‘realistic stimulation of sensors, replication of visual cues’ and ‘stimulation of more than one sensory capacity’.
Predeployment stress inoculation is one purpose of the ‘mixed-reality’ Infantry Immersion Trainer at Marine Corps Base Camp Pendeleton, and of the Future Immersive Training Environment at Camp Lejeune. The latter uses both head-mounted displays and suits equipped with sensors, enabling projected avatars to replicate the movements and actions of the participants. Rifle shots are wirelessly relayed to a receiver which incorporates the firing data into the virtual environment.
At these and similar sites, trainees are exposed to high-fidelity simulations of the ‘visual, auditory, haptic, and olfactory elements of the operational environment’: ‘sudden explosive impact, loud ambient noise, depiction of casualties, post-event chaos, presence of insurgents, and other stimuli’ including weapon signatures and ‘virtual characters [responding] realistically to being hit by small-arms fire.’
This is designed to produce an acute stress response from the infantrymen. Stress-response profiles are gathered by administering salivary-hormone tests that measure adrenal (cortisol) and alpha-amylase secretions, and by monitoring heart rates.
According to an evaluative report presented at I/ITSEC 2010, ‘exposure to realistic stressors may provide a degree of inoculation to the stress of the environment before it is experienced in theater, leading to better performance under stress’. This will, according to another I/ITSEC paper on the effect of training on mitigating stress, increase ‘the ability of forces to deliver the outputs for which they were designed’.
Such research retains the preoccupations described by former West Point psychology professor Dave Grossman in this blog’s previous post (i.e using repeated realistic drills to overcome the resistance towards killing of ‘warfighters’). But it translates the old behaviourist language into an up-to-date cognivitist account of how lethal behaviour can be learned then later elicited with ‘increased speed, increased accuracy, immunity to environmental stressors, and greater retention’.
The trick is to save (scarce) processing capacity by making automatic what would otherwise be a deliberate, consciously-willed action:
Controlled processes are those strategic processes required for novel tasks or those tasks requiring the devotion of attention. Controlled processes have been considered serial in nature and are carried out in a stepwise fashion. They have also been thought to be carried out much like a recipe, do this- then this, etc. Controlled processes have been described as being performed more slowly and under the explicit control of the individual. After much consistent practice, some tasks no longer require the devotion of attentional resources and are performed to completion in the presence of the initiating stimulus unless there is a conscious effort to inhibit them. Automatic processes are characterized by this decrease in the cognitive/attentional resources allocated for these tasks…An automatized motor-skill may operate at a level not associated with awareness, or not directly under conscious control…[Automatic] processing…is difficult to alter, to ignore, or to suppress once learned…An automatic sequence will tend to reoccur in response to the previously trained initiating stimuli.
[If] a person consistently chooses to pursue the same goal within a given situation, over time that goal structure becomes strongly paired with the features of that situation. After this consistent pairing of the situation with the intended goal, eventually, the goal itself is activated on the perception of features of the situation in a preconscious analysis…Performance improvements will occur in situations where stimuli are responded to in a consistent manner across exposures…Contextual cues should be used to mimic the effects of consistency and may activate automatic sequences of behavior…Training to levels of automaticity will also make performance reliable under environmental stressors such as alcohol, fatigue, heat, noise, and so on…In some cases, consistent practice can lead to direct memory access, such that the execution of the recipe or algorithm is no longer needed.
In short, ‘repeated experience is the best way to train decision-making skills so that a required cognitive process that might initially be analytic and labored can become more intuitive and automatic.’
UCF, alongside other institutions in the Team Orlando orbit (e.g. ARI and Lockheed Martin), has declared positive findings on the usefulness for this purpose of games and simulation. They make familiar to the trainee what would otherwise be novel and highly aversive sensory environments (i.e. all the perceptual information involved in killing another person in combat). And they allow behaviour modification via the same well-worn neuroendocrine pathways that are engaged during close-combat training and repetitive markmanship drills.
Visual-attention tasks are known to stimulate cholinergic release and thus allow rapid perceptual learning. People who engage in such repetitive, high-stimulus activities for extended periods are known to experience visual and auditory hallucinations that accurately replay the activity as they fall asleep (hypnagogic imagery, or the so-called ‘Tetris effect’). Such images are not limited to sleep onset but intrude during wakening; there is a lag between sensory stimulus and hallucination. This is known even to affect amnesics: those with deficits in conscious declarative memory.
Those kinds of training that elicit some instrumental motor response from the trainee or player (in a virtual environment, desktop computer, console or head-mounted display) are usually also designed for constant feedback and engagement of the user’s reward system, allowing proprioception and self-correction. In their most advanced form, such media gather precise activity data via click-stream logs, track sequences of decisions, detect anomalies or failure, and build behavioural profiles of the trainee or player.
Perhaps most importantly for training purposes, the user is desensitized by repeated exposure to high-fidelity depictions of physical violence against other humans: photorealistic vision of gruesome deaths, directional audio including ‘human distress noise’, scent machines, haptic feedback, wind, heat, and motor control over the action.
The boundary between processing of visual (and to a lesser extent auditory, tactile etc.) information and planning of movement is fuzzy. It seems likely that the primate brain’s comprehension of observed behaviour (say by an avatar) involves performing a ‘covert rehearsal’ of that activity in the parietal and premotor cortex. The evidence of facial imitation and automatic prereaching by human neonates shows a link between visual information and motor movements at the very outset of development, i.e. immediately after birth. But the learning of older children and adults also involves involuntary imitation and emulation of the perceived actions and gestures that are performed by those around them.
These are the biological correlates of the conditioning and social-learning process by which military recruits are made into technically proficient killers.
So, just as with the post-WW2 innovations in live marksmanship drills and operations training discussed in the previous post, some of the useful features of virtual battlefield simulations and videogame combat training (from the viewpoint of the instructors) are those that overcome a resistance to killing that otherwise incapacitates frontline ‘warfighters’.
But this raises an obvious point. Many or all of these features are also shared by popular visual-entertainment products: TV, movies, and interactive media including videogames.
These, taken as a whole and on any reasonable measure, expose users to an extraordinary barrage of realistically-depicted violence. Consequently their audience judges acts of violence against people to be less novel, aberrant and repulsive than they would do otherwise.
The ubiquity of salient and vivid mental pictures of aggression prompts people (by the availability heuristic) to overestimate the frequency of such events (in reality, though not at the cinema, homicide occurs less often than suicide). But this is a self-fulfilling judgement. If an individual fearfully supposes he is engaged in a perpetual Hawk-Dove game (a reasonable conclusion to draw from the entertainment media’s p0rn0graphy of violence), and that other players are likely to have chosen an aggressive strategy, then his best response too is to be aggressive rather than meek. The result is a breakdown in cooperation.
Of course, adopting such a strategy and executing it are two distinct things (there’s also a difference between signalling aggression, i.e. making credible threats of inflicting damage, and actually following through with violence).
The latter only occurs, as the above military research shows, when violence is made automatic and reflexive by instense and persistent brutalization. Such a programme can over-ride the social and natural barriers that otherwise keep conspecifics (including fellow humans) from killing each other, even when instrumental rationality tells them to do so.
Conflict between animals of the same species is commonly described as ‘limited war’, involving insufficient force, ritualized tactics, withholding of offensive weapons, and seldom causing serious injury to the participants. ‘Escalated fighting’ is generally absent from intraspecific battles, e.g. snakes merely wrestle each other.
But, to judge by skyrocketing rates of aggravated assault in most industrialized countries over recent decades, humans are trying to kill each other rather a lot more than they used to. All this is suboptimal, to say the least.
But is it just another regrettable case of market failure: a poor aggregate-welfare outcome resulting from individual firms each trying to make a buck, whichever way they can? Or does it serve some functional purpose?