Waiting for the Dot.Sim Boom

Simulation has not had its dot.com boom. Our field has always grown in lockstep with a number of high technology areas. We have ridden right along with, and in some cases have driven, the leading edge of new technologies. As mainframe computers became workstations, and then PC’s, simulation products leveraged all of this equipment and became better for it. As computer graphics moved out of the university research labs, we were quick to adapt these new technologies to our flight and driving simulators. We moved from large cabinet-sized image generators, to smaller graphics boxes, to dedicated workstations, to internal graphics cards. As networking and the Internet became prolific we connected our simulators together, created standard networking protocols, and constructed distributed events across all of our facilities.

We have always been very hungry for advanced technologies because our customers constantly demand more capability, better performance, and lower costs. Working in the simulation and simulator business has always given engineers the opportunity to apply the newest technologies emerging from research labs and commercial vendors.

But all of a sudden this stopped. The commercial world discovered that the Internet allowed them to do business in an entirely different way. They were able to connect directly to millions of customers around the world without creating physical stores and without shipping special equipment to every customer. The Internet opened the door to delivering products and services to every single customer in the world. With it the Amazon.com website could sell more books than Barnes & Noble with its 800 physical stores. This was a huge change in the relationship between a vendor and its customers. I grew up in a small town in Southeast Colorado with very limited access to retail products and professional services. A big shopping center was the Sears Catalog store. It contained a refrigerator, a dishwasher, and a table full of catalogs. If you needed appliances, lawn equipment, tools, or clothes you shopped for them in the catalog and placed your order. There were only a couple of storefronts on Main Street and not a single bookstore. My bookstore was a single small shelf at the local drugstore. My Amazon.com was the mail-in form on the back pages of the books I purchased from the drugstore. Barnes & Noble could not reach out to me with its vast selection. I had to reach into the inventory through the soda straw listed on one page of a paperback book.

Amazon.com is not just a huge warehouse of books. It is a delivery system that can reach every single person in the networked world. It allows a child in a small town of 500 people to access the same books as a child in the heart of New York City. It breaks down the location-specific barriers that prevent people from learning and exploring on their own initiative. That is the real power of the dot.com boom. In the simulation community we have not created this kind of service for our customers.

Simulation systems are still delivered like heavy products to specialized facilities. We create destination sites in the pattern of Disney World that soldiers have to visit physically to experience. And like Disney World, such visits can be once-in-a-lifetime experiences. Our soldiers cannot travel to our high-end destinations every time they want to improve their performance or explore a new idea. A dot.com boom in simulation would extend our systems through the military Internet to every soldier’s desktop computer. It would allow every soldier to browse our offering of simulation services, enter the one of their choice, and join a team to explore a new idea or receive a lesson from a leader.

The technologies to do this are available now and pooling at our feet. But we continue to insist that training via simulation requires a dedicated facility, specialized equipment, and a large support staff. We insist that simulation cannot do a soldier any good unless it is custom crafted by an experienced professional and makes scant use of the newest technologies. To continue the Amazon.com analogy, we are insisting that books should only be sold in physical stores by a trained staff, and that an online bookstore would corrupt this process by allowing people to select their own books without explicit human guidance. This same argument was made against online education for years. Online universities were once considered the lowest form of crass commercialization of a much higher calling, not much above a diploma mill. But today, every university from Harvard to the local Community College offers some or all of their degree programs on the Internet.

The technical tools already exist to provide Internet-delivered, simulation-driven, training and exploration. What does not exist is the will to customize and extend our resources to reach every soldier in the service. We still want the soldiers to come to our specialized facilities and our dedicated staff. We are not ready to let soldiers take a hand in guiding their own training.

We currently have wargames that can be adapted to run on the server side of these training networks. We have computer games that can provide the intuitive client side interface. We have IT infrastructure tools that can tie the right soldiers to the right applications. New “Web 2.0” applications are opening the door to user created and modified content like our simulation scenario databases. These will allow a soldier to modify existing scenarios and select specific AAR products from within a standard browser. User interfaces like Google’s Earth and Map products can provide a window into a simulation running in the military computer cloud. Though DOD security regulations may make the use of such tools difficult, they do not make it impossible. What is really difficult is to get our community to see simulation as a service that can be extended to millions of soldiers, rather than as a device, a facility, or a destination experience.

Each time our industry grasps new technologies there is the fear that we might abandon the good work that is being done with the older methods. But these fears are never realized. There was a time when all training was Live. The emergence of board wargames provided tools to supplement and extend that live training, but not to replace it. The computer revolution brought us virtual flight and vehicle simulators, but these did not replace live training. Instead they allowed pilots and vehicle crews to experience situations that were impossible to create in the live world. Simulation as a Service will expand and extend the value of our products to the soldier in the same way that virtual and constructive systems have done in the past. Some tasks can be trained very effectively from a standard desktop computer. Other skills must remain in the live, virtual, or constructive systems that we already have.

A dot.com boom in simulation will do the same for the soldier that Amazon.com did for the reading children of small town America. It will make huge volumes of training material accessible to every soldier, everywhere, all the time.

History of Games: From Dice to Computers

I have created a presentation on the History of Games. It was first presented at a defense simulation & gaming conference so has definite bias in that direction. But the long-term goal is to pull together a sahred resource that many people can use in education.

You will find the PowerPoint on SlideShare.net and are welcome to use it for your own non-commercial uses. If you make significant additions, please send them to me so I can incorporate them into the briefing for other people as well.

History of Games: From Dice to Computers

View more presentations from Roger Smityh.

Games vs. the Wizards of Grep

Games are making a big splash in the military training world. They offer beautiful graphics, simplified user interfaces, and low cost to entry. But some people question whether they are really valuable simulations. They point to the lack of detail in some models and the singular focus on direct fire combat engagements. But that is the same way that a lot of simulation have entered the inventory. One of the earliest models, CARMONETTE, was introduced in 1953. The first version contained only tank-on-tank combat with a little anti-tank play. Each year new capabilities were added - infantry in version 2, helicopters in version 3, and communications in version 4.

Games are going to evolve like that as well. They are being introduced in their sweet spot. Once they are running well in that specialty, they will grow into other areas. We will discover how to apply them to situations that are unique to the military and that have not be perfected in the commercial entertainment space. Such perfection in the entertainment world requires 100's of millions of dollars and many years of trail-and-error product launches. They are not going to do all of that work for functions that draw in 10 or 100 customers.

Remember when all computers were used from the command line? We all learned cryptic commands like ls, grep, chdir, chpwd, and hundreds more. With these and their endless list of arguments we could invoke thousands of different actions on the computer. But then we were presented with X-Windows, the Mac interface, and Windows. These hid the magic commands and allowed anyone to hunt-and-peck their way through a set of menus to invoke the same actions. The Wizards of Grep all proclaimed, "too slow", "for amateurs", "not flexible", "can't really see what is happening". But these interfaces opened the door to millions of new users of computers. They allowed people to use the machines without becoming wizards and they eliminated a lot of the tragic mistakes that even experts made. (Remember your first accidental "rm -rf" command? ...oops...) This led to computers as artistic devices, multi-media machines, and replacements for all office equipment.

Games will do for simulation, what windows and Windows have done for desktop computing. They are also undeniable and unstoppable.

Games vs. Virtual Worlds for Nation-sized Problems

3D Shooters are the most prominent form of game system and environment in the consumer and the defense space. These portray conflict, combat, and deadly threats. They immediately plunge the player into a simulating environment with urgent problems to solve. They also mirror some of the most important engagements that real people and real societies engage in. However, these environments are extremely limited in time and space. The battlefield is a relatively small area – usually just large enough to contain a specific vignette, and never so large that the players can wander far enough to miss the entire point of that piece of the world. These vignettes and geospaces are linked together in such a way that the player can move immediately from one “hot spot” to the next. There is no room in these for intervening relationship building, downtime learning, AAR, or planning for the next engagement. For entertainment this hot-spot-hopping is exactly what you want. But as a venue for wrestling with real problems, this is a very small and single-focused experience.

MMOGs create a much larger space in which player spend more time wandering, conversing, building relationships, and joining clans that will participate in specific battles. It includes spaces for combat, socialization, trade, and exploration. This size and diversity enables a much broader and somewhat richer experience of the world and the other players in it. Specific battles may still be the focus for many players, but they can also plan, rehearse, and regale in stories surrounding these as well. The algorithms that determine engagement outcomes, but battle and trade, are simple – often just subtracting and adding points to a player’s health.

Virtual Worlds create an world that can be smooth and continuous like the real world. They can create context, connections, and history that is similar to what exists in the real world. But in their current state they are only slightly different from MMOGs. Second Life, and others like it, are unique in that the content is created by the users, not by the development company. This begins to allow the users to shape the world to meet their needs. But to really become distinct and useful, these spaces need to allow the users to upload/link their own models into the world. The VW needs to provide an infrastructure that can accommodate heterogeneous models provided by users and allow these diverse models to interact with each other. Business and Government problems cannot be represented by generic one-size-fits-all models provided by an entertainment company.

Each game designs a set of models that meet the needs of that game. The preference is to create sparse models that are computationally inexpensive and that fit together to allow interactions across all of the objects in a space. As virtual worlds are adopted to the needs of real government and intelligence customers, there is going to be a need to (1) add much more complex models that require more computational power, (2) bring together a very diverse set of models that were not originally meant to work together. A government Virtual World cannot align these models one at a time, that is an N-squared problem that will very quickly become impossible to manage. There needs to be an infrastructure that allows heterogeneous models to be integrated into the world and to work with the existing models without requiring customer model-to-model modifications. This would be a big environment with an underlying software infrastructure that present real value to the government.

For the complete briefing see: http://www.peostri.army.mil/CTO

2008 Defense GameTech Conference

The agenda and presentations for the 2008 Defense GameTech Conference are posted on the web at:
http://www.peostri.army.mil/PAO/pressrelease/20080425_gametech.jsp

[Note: This is a belated posting. Google does not seem to have indexed the link above.]

The 3D Cellphone

Big computer companies like IBM and HP are developing software that allows you do view highly detailed 3D worlds on modest client machines (see DCV and RGS links below). This is accomplished by doing the rendering on the server side and sending the screen to a number of smaller clients.

Companies like Vollee and OTOY are doing something similar, but targeted at the cellphone. Imagine that you are running a 3D world like Second Life on your cellphone. But, since your phone does not have the compute or storage resources to really do this, all of the rendering is being done on a server and the results streamed to your phone in the form of a digital movie (MP3, Flash video, etc.). Your inputs on the phone are commands to move through the virtual world and interact with the objects there. These commands are carried to a server where the simulation and graphic rendering are done and the finished video frames are streamed back to your phone for you to see. Clearly there will be some video lag between the command and the visual results as the key entries travel to the server, are simulated and rendered, and the results travel back to your phone. if you are old enough you will remember that this is how text entry and order execution worked with the old terminal windows back when "the Internet" meant textual applications on a command line and there was no such thing as "the Web". You may also have seen the gradual evolution of that primitive interface into rudimentary graphic menues as clever people showed that the text could acrually drive a menu system rather than just showing up on a command line.

High-def rendering on all devices will become a reality. A few years ago we thought it would happen through the miniaturization of the GPU so that all phones had an Nvidia or ATI chip in them. But faster networks are making this possible while the GPU remains on the server machine. The connection between the client and server is fast enough that the two seem to share the visualization capability. There may be several technical hurdles to work through, but the community have solved bigger problems than those to get where we are now. Keep looking for great things on small devices.

IBM: http://www-03.ibm.com/systems/deepcomputing/visualization/
HP: http://h20331.www2.hp.com/hpsub/cache/286504-0-0-225-121.html
Vollee: http://www.vollee.com/secondlife
OTOY: http://www.techcrunch.com/2008/08/20/the-truth-behind-liveplaces-photo-realistic-3d-world-and-otoys-rendering-engine/

Virtual tea parties on the ceiling

Why do we have chairs in our offices, classrooms, and public spaces? Are they a real necessity? Somewhere in the distant past people recognized the advantage of doing certain types of work mounted on a platform up off of the ground. They also found the need to keep nice clothes off of the ground and out of the dirt. Then there is the comfort factor for sitting down rather than standing up. In the physical world, chairs have become a staple of our existence.

When we enter the virtual world, what is a chair good for? There is no dirt, there is no physical fatigue, and characters can often float anywhere? So what use is there for chairs?

But if you navigate through Second Life you will find chairs, desks, lamps, and all of the typical artifacts of the the real world. All of them useless. We are so excited about the possibilities that can be achieved in the virtual world. But once inside it appears that most people cannot imagine anything different from what they have in the physical world. In fact, it looks like the virtual world is just a place where we can possess something that resembles what we cannot get in the physical world. If, in the physical world, you have a small office or cubicle, the first thing you might build in the VW is a bigger office. One bigger than the boss'.

Where is the imagination? It appears that most VW residents have little and most large organizations that enter the VW have absolutely none. Even though people complain that sitting in conference rooms and classrooms all day long is the most boring part of their lives, that is the first thing we offer them when they enter the virtual world.

Given an empty virtual canvas we create images directly from the physical world. We create load-bearing columns for buildings that weigh nothing. We hang light fixtures from virtual chains even though there is no gravity to pull them down. We put all of the furniture on the "floor" and leave the walls and ceiling of buildings empty. Where are the tea parties on the ceiling?

Second Life has many very clever and imaginative residents. But for some reason the islands of large organizations are just as sterile, standard, and unimaginative as the real offices. The "rules of the office" carry into the virtual world, even the rules that call for chairs that serve no purpose.