Col



WARGAMING in 2015

Lt Col (s) Peter Garretson

Address to the Air War College Class 2015

Col. Gwen Ali’i

It is an unusual honor in this day and age of time-shifted media for an Air War College class to request a real-time net-meeting, so I will try to be worthy of your request and give you a useful overview of the methodology the USAF follows in its Future Capabilities Wargaming system.

Both the ends and means of the Future Capability Wargame has evolved since it was introduced by Gen Ron Fogleman in 1994. While we still use the game to discover insights about the utility and employment of new weapons system concepts, to probe the nature of future conflict, and to help inform planning decisions for force structure, we now use the game to do so much more.

As one example, we no longer sit down and try to make up an interesting and credible scenario that provides the “blue-preferred red threat” prior to a given wargame. These days, we are interested in actual insight into where we might actually become engaged—in whatever year they may fall. We now have several tools in our arsenal to examine these.

The first is the Air Force Conflict Decision Prediction Market.[i] This artificial market harnesses the “wisdom of crowds” to determine via betting where conflicts are more likely.[ii] Another is via the USAF open submissions program, where anyone in the USAF stakeholder community can submit a problem or scenario, and much like in the old Digg or Slashdot websites, user ranking bubble the more worthy to the top. Many of our best leads come out of AFRL’s Focused Long Term Challenges (FLTC’s)[iii] the continuing Blue Horizon’s[iv] future studies programs started a decade ago to systematically look at technology and future strategic environments, and the Red and Blue “Discovery Games” that help flesh out and prioritize possible Red and Blue capabilities.

The third is the implications wheel.[v] In this case, when one of our senior leaders wants to explore the potential implications of a significant decision, they rapidly fire off the decision to a large and diverse number of individuals who rapidly capture and score multiple cascading downrange implications, both positive and negative to appreciate their effects and prevent unanticipated blowback.

Fourth, we have leveraged complex physical & climate models from NASA and NOAA[vi] with economic and demographic models from the intel community[vii] to provide a backdrop of trends stressors.

On top of this, we play a turn-based National Decisionmaking[viii] / POM Game for all 193 countries from the present forward some 50 years. To do this, we rely on a cadre of about 2,000 poli-sci students at various universities and ROTC programs who donate an occasional weekend to play various contingencies within countries and regional players who must achieve national goals and stay in power domestically. They do this via collaborative software we provide. This was actually the first part of our wargaming system to become distributed. When, in the course of their play they encounter a crisis or conflict, it is forwarded for military wargaming, complete with a “history of the future.” Along the way, we gain a number is key insights about regional dynamics, and likely trouble spots that inform our force structure and basing decisions, and help us advise our national authorities on how best to use USAF posture and force structure for Grand Strategy. We are currently in the process of evaluating various different artificial intelligence engines in various roles in this game so we can run higher volumes of different scenarios.

Once we have adequate “scene-setters” and a library of scenarios, the next step is to play actual wargames in those alternate futures.

And here too, our actual game play has significantly evolved. Of course, the eye-candy has gotten very sophisticated—a decade of progress in consumer gaming and AFRL FLTC 1[ix] has provides us all with superb low cost augmented, mixed and virtual reality that both enhance gameplay and help us communicate strategic concepts through their cinematic qualities as machinima[x], but that is just the outer-most layer of the onion, and the true progress goes much deeper.

From its origins as a Red-on-Blue turn-based BOGSAT[xi] where generals put down their thoughts on PowerPoint slides and the occasional after-the-fact aggregate modeling, we have moved to increasingly sophisticated mimics of the actual C2 environment incorporating both turn-based and real time strategy elements against a persistent, virtual, entity-level world.

But about ten years ago game developers debut massive multiplayer online games[xii], persistent virtual spaces[xiii], sophisticated entity-level turn-based and real-time strategy games[xiv], and commercial grid-computing[xv]. DARPA was also working on sophisticated mission preview[xvi], cultural learning, command decision support tools[xvii], and artificial intelligence both for blue[xviii] and red[xix] entities, while AFOSR was doing cultural modeling[xx] and AFRL was creating synthetic warriors[xxi] and new physics based models. All these capabilities developed very fast, to where we are today where we are able to run multiple, concurrent, theater size & entity level persistent virtual conflicts on demand, and players can incarnate as any entity at any level.

Now that would have been impressive enough when it was first articulated in 2006 under the USAF Strategic Plan for Modeling Simulation and Decision Support, but what really made this so powerful was the ability to deploy it to each and every member of the USAF along with very powerful user-level tools and wizards to configure and run a sim in a matter of minutes.

And of course it is agile and low cost to maintain because most of the software is free and government owned. Of course it routine for us now, but back then the idea of relying on the patriotism of our citizens to do sophisticated software coding Open Source[xxii] without monetary reward was a radical paradigm shift for a system used to Cost-Plus acquisition. There are of course certain high priority updates for which we have to offer prizes like an F-35 ride, but most routine changes just seem to happen on their own once we released the core engine as “America’s Air Force.”

We owe the tremendous (and ever increasing) utility of these PC-level sims to the early users and architects who were adamant that the system be completely modular. By separating the user interface, graphics engine, rule & adjudication engine, artificial intelligence control, each could follow its own natural evolution, and the software could be used at any security level with “drag & drop” entity, rule, and geospatial databases.

And those databases have not stood still either. When efforts began, the Toolkit for the Futures Game consisted basically of a cartoonish database of single page descriptions. There was no single source authoritative, multi-level entity descriptions, models and force structure for both red and blue. The data was everywhere--A8XC, A5XS, A3OC, A9, NASIC—split between intel, analysis, wargaming, and R&D, and not adequately accessible or maintained. What we did was to follow the community model pioneered by Wikipedia, and bridging on existing efforts of CISAN and TMAP multi-level security and Intellipedia, we added value to the community by creating a community-policed Wiki for each system, system of system, and force structure in a community space where attached to each are the latest models for all applicable sims. SAF/XC and AFAMS maintains the pipes and the format, but the entire content is self-policing—changes are tracked and attributed, and each system is appropriately tagged or approved as each organization gives its blessing for particular concepts or games.

On the geospatial side, it is a rare game that needs anything other than publicly available open source. When these efforts began, Google Earth was still a novelty and only displayed historical aerial and satellite images at about 1 meter accuracy, now of course it integrates real-time feeds from sources as diverse as commercial satellite and UAVs to individual webcams and cell phones overlaid on consumer/hobbyist LIDAR models. Various computer vision web services also process this data as well as state-run Air Traffic Control Radar and hobbiest multistatic radar to provide real-time global Moving Target Indicator (MTI)[xxiii].

Back then, those iPods you all carry had a maximum of 60 GIGs, whereas today, with storage having doubled every 18 months, those same iPods can now store 64 times as much, all of 38,480 Gigs--the entire land area of Iraq at 64 bytes per square meter.[xxiv]

But I digress. What this proliferation of tools meant, was that suddenly we had the tools to do entirely new things. The biggest innovations, have not been the technology itself, but how we organize and have learned to employ it.

For the first time, if someone—anyone--had an idea of how a battle could have been conducted better they could change the tactics or force structure or introduce a new notional tool and have an objective basis for comparison against an ever-growing standard library of historical and constructed military operations. As new ideas prove successful, they are tried out with other strategies in a sort of genetic algorithm.

This system has become quite important both in acquisition and concept development—to include the design and testing of the actual C2 interfaces and collaborative tools. Because of the grid server scalability (which is inexpensive and commercially purchased by the CPU-minute), once a general concept is identified as useful, the exact optimum attributes and force mix can be discovered through multiple monte-carlo analysis and genetic algorithms[xxv] that mix and re-introduce the best elements for multiple match elimination. We are also now able to make use of spare CPU cycles on government and donated computing time because we are now configured to farm out our analysis in Grid computing using screen-saver spare CPU time.[xxvi]

As many of you know, up until this last decade, the USAF, like most militaries grew and selected its operational level strategists and commanders based on tactical performance, largely because there was no other more reliable way to identify talent.

For the first time we had a high fidelity simulation for command and control functions which could provide synthetic experience. That meant that we could now train and select for commanders based on actual talent in those specific skills with fidelity never before possible.

With respect to selection, it meant that we could select people earlier and earlier, the America’s Air Force Game[xxvii] has been a watershed. Not only has it been an exceptional public relations and general recruiting tool, but it helps us identify both talented future strategists and tacticians as well as harvest innovative tactics and strategies themselves.

With respect to training and experimentation, would be commanders could begin training and experimentation very early in their career, and has allowed us to significantly raise the bar. As your readings have no-doubt made clear, the USAF developed various metrics to be maximized and provides an on-going challenge and competition to generate strategies and tactics to achieve a maximum score.

Some of you are already familiar with the games. For the rest of you, I can certainly promise that your next few weeks will be very engaging and likely mean some sleepless nights. The scenarios that are part of your curriculum were grown from the forecast system already mentioned, and then chosen by senior generals via internet voting to determine the most relevant.

At each level of PME, we tailor the games to allow you to play in the world and with the force structures you are likely to encounter when you actually enter command. Think of the individual games as a sort of Red Flag metaphor for command—the first few sorties to ensure you’ll “survive.” In some cases, you will even be allowed to design your own force structure, so you get an appreciation of the difficulty of making the trades and how the POM game is played. For the most part, you will play the roles appropriate to your career track and ambitions and primarily interact with sophisticated AI’s that mimic the actual people in the CAOC, but you will also get a chance to play other roles to gain an appreciation for their problems and considerations. You may also find that since your games are entity level and take place in a persistent virtual space, other training audiences may choose to incarnate / spawn in place of the AI’s through any number of the virtual flag squadrons that use these exercises as backdrops for their own training and exercises. First you’ll cut your teeth just you against the computer, but later you will get to play each other, faculty members or well known strategists & mentors.

As we have increasingly automated our AFT2EA process with computer based tools, farmed out human intelligence “microwork”[xxviii] and rely increasingly on at-home citizen-analysts for low level work, we are able to free-up intel talent from broad area search and targeteering to more strategic and value based analysis. As we continue to rebuild our Red Team capabilities and expertise, Red will be increasingly played by our professional country cells by intel cadre. When entities are automated, they often reflect tactics codified in more tactically focused virtual flags and distributed mission simulations.

All games can be used for experimentation and may be posted anonymously. However, for those of you who are ambitious or desire attention, I can tell you that GO’s do watch the progress of posted games and regularly attend to the highest scoring games. Many jobs and some commands do request certain game scores and metrics when competing for jobs, and the criteria vary greatly, as the metrics measure such things as risk-taking, innovative thought as well as time to completion, cost, blue and enemy losses, exposure to threat, etc. New metrics are always being evolved and posted as well.

However, as engaging as the single player strategy games are, I suspect you will find the multi-player group games the most interesting. This entire part of the curriculum is based on a very different theory than AF futures wargaming in the past. The USAF’s own strategist, Col John R. Boyd’s articulated a theory of organic command and control that stressed the importance of implicit over explicit communication, which ensures security and low bandwidth, and enables harmony/rapidity/variety through a shared orientation. Boyd felt the best way create this was to allow the different people with different skills to interact with each other and the environment to generate shared images and shared understanding—to know exactly how your colleagues are likely to react in a diverse set of circumstances. So these games are picked to provide a number and variety to provide a broad toolkit of skills—think of it as a kind of mental Leadership Reaction Course. Of course you will get to play a teamed roles in a difficult real-time MCO, and you will get to mobilize for other contingencies as well. Like the old adage that “if the only tool you have is a hammer, every problem looks like a nail,” we don’t want you to be armed with only a single paradigm of Airpower, since the nature of Air, Space, and Cyber Warfare may rapidly change in your career timeframe and require alternate strategies and maverick thoughts rather than get trapped in “old-think” or “obsoledge.”

We have deliberately crafted the individual scenarios[xxix] because they challenge normally safe assumptions, force you to think the unthinkable, and create a rich tool set—but they will require a deliberate suspension of disbelief. Again, realize that just like the purpose of lifting weights is not to lift the weights but to build up the body, the purpose of these games is not really to plan for these far out contingencies, but to inspire cohesion and implicit communication.

In summary, the USAF Future Capabilities Wargame has evolved over the last a decade from a single, in-person MCO scenario for a small audience of General Officers recorded on PowerPoint to a network of distributed games integrated into our Professional Military Education--designed to realize an organic theory of command and control by creating trust and harmony between our officers and exercise their mental agility and cooperative and competitive skills. We have gone from individually concocted scenarios to re-enforce a dominant MCO theme with other operations considered lesser included cases, to a proactive scenario forecasting capability generating a broad range of possible scenarios. From a single BOGSAT event, we have moved to an on-going relay of strategic and tactical exploration and sharpening of our command and strategy cells. And we have moved from a system that existed in isolation, to one that directly answers strategic questions and helps us select the most fit for actual strategy and command. The strategic center of the wargame is no longer a paperweight report delivered through a snail-mail system, but the deliberate sharpening of tactical and strategic thought and the creation of shared experiences that create the trust and shared orientation for implicit rather than explicit understanding, where a single short phrase is worth volumes of guidance, not unlike having an “inside joke” or a twin’s private language.

Our role in A8XC has also evolved from hosts and scenario developers to our new role as project leads for open-source content development and selection and to aggregate and provide insight. We also adjust the various incentive systems to ensure adequate attention is being paid by the players to examine the “elephants in the room,” challenge our cherished assumptions, explore the strategically relevant “should the USAF turn left or right” questions, discover downstream implications and possible blowback to them, and try various adaptive approaches to better characterize the pros and cons of strategies to attack our “Wicked Problems.”

-----------------------

[i] Decision or prediction market,

[ii] The Wisdom of Crowds by James Surowiecki

[iii] FLTC’s were began in 2005 and provide mid and long term demos of various key persistent problem areas to bring about the AF tech vision of AF2T2EA4

[iv] Blue Horizons began formally in 2006 at Air University in partnership with AF/A8XC, and is overseen by the Air Force Futures Group (AFFG), consisting of about 12 members from various parts of the USAF doing long range thinking and planning

[v] Joel Barker, the Implication Wheel or i-Wheel,

[vi] NOAA Model NASA Model

[vii] and

[viii] National Security Decisonmaking Game (NDSM)

[ix] FLTC 1 contains a subset called “Continuous Learning” which aims to create an immersive distributed learning environment

[x] Using computer games to create movies is called Macinima for example, Red vs Blue:

[xi] BOGSAT = Bunch of Guys Sitting Around Talking

[xii] MMOG , examples: PlanetSide , World of Warcraft

[xiii] Persistent Virtual Space or World , examples include OLIVE , There , and Second Life

[xiv] Turn Based Strategy Games for example Civilization IV , A Force More Powerful , Real Time Strategy Games for example Command and Conquer , Caesar IV , and SimCity

[xv] Example, Amazon’s Elastic Compute Cloud

[xvi] DARPA Rapid Mission Preview program, a.k.a. “RealWorld.”

[xvii] DARPA, numerous, including Integrated Battle Command (IBC), Command Post of the Future (CPOF) , JAGUAR

[xviii] DARPA training superiority DARWARS program

[xix] DARPA Realtime Adversarial Decisionmaking (RAID)

[xx] AFOSR Cultural Modeling Program casos.cs.cmu.edu/bios/carley/KC_vita_March_06(4).doc

[xxi] AFRL Combat Automation Requirements Test Bed (CART)

Synthetic Warriors

[xxii] Open Source Movement ,

[xxiii] Free Moving Target indicators, such as: for satellites, or or for airplanes (or see their route in Google Earth: ) or use of cell phones to track traffic flow: or keep track of children:

[xxiv] Memory doubles every 18 months, that is 8 doublings between 2006 and 2015: 60 – 120 – 240 – 480 – 960 – 1920 - 34,480 Gb = 34,480,073,741,824 bytes / 437,072 sq km in Iraq = 78,888,772.9 bytes / km2 = 78 bytes per m2. [To store 1 km x 1 km of data at 64 bits per mm2 = 6.4 × 1013 bits or 8.0 × 1012 Bytes, your IPOD could store 4.3 sq km of area at 1mm resolution]

64 bits per sqr mm pixel = 4,310,009,218,750 Sqr mm of data = 2,076,056 mm on a side area.

[xxv]

[xxvi] which runs Grid computing for Climate Models, Cryptography, Protein Folding, and SETI.

[xxvii] While both the Army and Marines have games, currently no America’s Air Force exists or is in any plans. A step in the right direction is AETC’s Tip of the Spear program to teach Expeditionary Air Base Opening, and ACSC’s use of John Tiller’s Modern Air Combat—neither of which are for external audiences.

[xxviii] For example, Amazon’s Mechanical Turk farms out small human recognition tasks as if calling a subroutine:

[xxix] Example Scenarios: “How will you adapt your organization (USAF) to the following new realities:

- “The US declares it is policy that the world will not go over 500ppm CO2 in atmosphere”

- “A fourth devastating hurricane rips through the gulf, there is scientific consensus this is a result of global warming, and the USAF is given the role of hurricane intervention.”

- “The Army submits a formal package to congress to dis-establish and cannibalize the USAF.”

- “Asteroid Aphosis is determined to strike the Earth in 2029 and crash in the mid Atlantic, devastating American and European Coasts.”

- “The SETI observatory receives an unmistakable signal from another civilization 8 light-years away.”

- “A million Americans are killed when a Japanese Eco-Terrorist sends several micro air vehicles (MAVs) to San Francisco with a BioWeapon.”

- “North Korea launches a Taepo Dong missile test that lands one mile outside US territorial waters (international waters) near Seattle.”

- “Terrorist Cells using open source deliberately attack our mobility assets on the ground.”

- “Peak Oil is declared by the DOE and Exxon Mobile, oil is over $100/barrel, and is expected to be unavailable at any price within a decade.”

- “Hizbollah electronically highlights an Israeli Drone and turns it against TelAviv.”

- “In a squabble, Iraq uses cyberwar to shut down key oil refining facilities and nuclear power plants in a neighbor.”

- “The Chinese beat us to the Moon and declare sovereign territory over a significant portion.”

- “A private multi-national firm demonstrates its ability to move an asteroid into Earth Orbit.”

- “Dubai’s space tourism port takes off, Congress decides the US cannot be left behind.”

- “The US stops a petroleum tanker headed for China under (erroneous) suspicion it is carrying WMD materials causing a blackout in the tight Chinese energy market. China declares US ships are no longer welcome in the Indian Ocean, and begins tailing with their subs.”

- “A Russian inventor demonstrates the ability to use a gravity force beam to bump the inertial system of one of our satellites.”

- “A Chinese laser lab succeeds in generating high frequency gravity waves allowing transmission and imaging line-of-sight through the Earth.”

- “The Federation of Concerned Global Scientists sets up Global Mil Watch, a series of mirrored websites to present real-time fused data showing the location and nature of every singe military entity.”

- “In an African peacekeeping mission, in a single day, we loose a HALE and two F-22’s—examination of wreckage suggest they were attacked by a prototype Directed Energy Air Defense.”

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download