Simulator Networking or SIMNET, which began with a young scientist's idea, ended up changing an entire industry and the way the military does business. The story is not over yet. SIMNET was an advanced research project aimed at the development of a core technology for networking hundreds of affordable simulators world-wide in real-time to practice joint collective war fighting skills and better acquisition practices. It was a daring project that proved the Advanced Research Project Agency (ARPA) mission of doing what previously could not be done. It was a serious threat to the existing simulation industry that resulted in added headaches for the government program manager. As it turned out, the government got what it wanted--a low-cost, high-performance virtual simulation capability that could be proliferated like consumer electronics. This paper provides an insider's view of the program history, identifies some possible lessons for future developers and opines future growth for the SIMNET technology.
INTRODUCTION
Numerous detailed histories have been written about the ARPA SIMNET program. It is not this author's intent to repeat those good works. Rather, it is hoped that a perspective of one who worked on the inside will be of interest to future developers. Hard lessons were learned. Old habits die slowly. In the end, courageous decision-makers won the day. Dr. Victor Reis, former Director of ARPA, used to say that, " the job of ARPA was to do what can not be done." In the early 1980s, it was generally recognized that it was impossible to build an affordable, large scale, collective, free-play, force-on-force, interactive, worldwide networked warfighting system. Fortunately, there were a few who accepted Reis' challenge. Because SIMNET comes as close to a revolution as any ARPA technology in recent memory. It not only changed the way the military does business, it also changed the simulation industry. This paper attempts to get at the substance of what was done and how it has influenced the future.
THORPE'S VISION
On 15 September 1987, a young scientist named Captain Jack A. Thorpe, with the Air Force Office of Scientific Research at Bolling Air Force Base in Washington DC, wrote a paper entitled, "Future Views: Aircrew Training 1980-2000." Thorpe hypothesized that, "... advances which are seen on the horizon are not simple improvements in teaching techniques or higher fidelity simulators, but rather bold concepts which tightly align training systems with real combat readiness and make them indistinguishable." Thorpe's original concept is depicted in the following four scenes.
Scene one (below) shows a real time overhead source collecting information about simulations worldwide and communicating these data to distributed simulation planning centers.
Scene two (above) shows planners analyzing the situation using a three dimensional holographic rendering of the denied area with analysts planning a future operation.
Scene three (above) shows four aircraft simulators in which crews are flying in the virtual environment and executing the plan.
Scene four shows the chain of command watching the real time dress rehearsal and assessing the overall plan.
A SHORT PROGRAM SUMMARY
In 1981, Thorpe was assigned to the Defense Advanced Research Projects Agency (DARPA now called ARPA). In 1983 Thorpe, with the help of Dr. Craig Fields, initiated Simulator Networking (SIMNET) with the goal of developing a new generation of high tech, realistic, networkable, microprocessor-based simulators that cost 100 times less than existing simulators. DARPA teamed with the Army to demonstrate the goal in a combined arms environment, which eventually ended with a network of 260 simulators at 11 sites in the United States and Europe. In 1990, the program was transferred to the Army and in 1992 the Army began a procurement under Colonel James Shiflett, Program Manager of Combined Arms Tactical Training System (CATT). In short, over a 10-year period DARPA and the Army invested approximately $300 million to develop and prove the SIMNET technology. Now the Army is in the early stages of committing roughly $1 billion to acquire a global 21st Century large-scale network of virtual simulators for collective training and combat development.
TWO LEARNING ENTREPRENEURS
The success of SIMNET cannot be discussed without the full recognition that Jack Thorpe was the dominant force behind its success. He spent a dozen years in ARPA during which time he took networked simulation from concept to reality. He had the technical expertise, tolerance for high risk, high personal energy and a sharp eye on the long-term goal. It is more than passing interest that at about the same time that Thorpe was thinking about his concept that became SIMNET, another visionary was working along the same lines. General Paul F. Gorman, while assigned to the Training and Doctrine command (TRADOC), envisioned a worldwide networked training system that connected field forces with the Army schools. The idea was to create a distributed learning system of relevant subject matter experts as tutors for on-the-job operators deployed around the world. In 1983 Thorpe and Gorman met for the first time in the Pentagon to discuss potential uses of the low-cost computer image generator that ARPA had under development at Boeing Aircraft. Later in 1985 Gorman and Thorpe became close collaborators in the development of SIMNET and its successor applications.
The pure serendipity of this meeting brought together two of the military's most creative and prolific training minds. Cumulatively, these two uniformed officers conceived and delivered profound learning systems for the military during their careers. Independently and later together, they worked to advance two of the most powerful overarching models in warfare:
Model # 1 says that commanders must prepare joint forces (Read: Units) for war by focusing at the point of the arrow. The preparation must be done with the tactical, operational and strategic commanders actively participating together in an experiential learning environment. In the 1970's Gorman developed the Multiple Integrated Laser Engagement System (MILES) for this purpose. In the 1980's, Thorpe did the same with SIMNET. In the 1990's, they advocate tieing them together in the Synthetic Theater of War (STOW) (another DARPA concept) along with wargames.
FOCUS ON THE POINT OF THE ARROW CHART
Model # 2 says that if we must fight we must fight with firepower, not manpower. Since the Civil War this has indeed been the trend. This chart shows the number of people per kilometer of battlefront over the past 200 years. The challenge is to expose even fewer people to the hazards of war but to arm each person with greater amounts of firepower. And most importantly, prepare (Read: Train) each person at all echelons to adroitly apply the greater firepower.
See Scene 6 (above) as the FIGHT WITH FIREPOWER NOT MANPOWER CHART
CONVERGING TRENDS
Early on Thorpe saw three dominant trends that could be merged to accomplish his goal: (1) the needs of the warrior, (2) advanced technology and (3) disciplined training requirements.
The first trend was the clear need for a capability to practice critical combat skills which could not be practiced in actual aircraft by in-flight training nor via substitution with flight simulators. The visionary warfighters of the force were demanding a greater capacity to get at the skills that were essential for survival and winning in combat. So the "Aces" were on Thorpe's side. The analysis of the required level of proficiency verses the capability for training at all echelons shows the disparity at unit levels of command. SIMNET was designed to fill that training shortfall.
Second, rapid advances in the fields of computers, communications and display technology along with their reduction in cost would facilitate his idea of networking large numbers of low-cost simulators together. These advances made it both technically possible and affordable. Here (above Scene 7) is the dramatic chart that ARPA likes to use showing the increasing computational power of over time. Both Thorpe and Fields believed this chart.
See Scene 8 for the Reis chart
Third, a behavioral discipline was coming into being that structured collective training skills based on realistic, measurable training requirements. Training analysis had previously been applied predominantly to individual tasks, duties and jobs. In the late 70's the military services signed an agreement to apply the techniques of Instructional Systems Design (ISD) to their training management. This let commanders and training managers focus their collective task analysis on the skill deficiencies that were caused by the inability to practice and thus selectively design the functional fidelity of the simulation to save resources by not producing superfluous knobs, dials, controls and displays. Said differently, this approach disciplined the developers into producing a simulator that could be affordable in large numbers.
THE POWER OF AN IDEA
SIMNET is not a person, place or thing nor system. It is an idea. Like the idea of the American West it is constantly changing. In the early beginning of this country, the West was Eastern Kentucky. Later the West moved to the Mississippi River and then on to California and now somewhere around the Rockies. The idea of SIMNET is no easier to pin down. In the late 1980s, SIMNET meant high risk to most of the military researchers and financial threat to the standard simulation makers. Then, in the early 1990s, it meant opportunity, a chance to be part of the revolution and maybe make some money. Today, SIMNET has come to mean what it was all along, a new technology that is the basis for application to many challenges. It always meant excitement. But like the genius of the American West, the genius of SIMNET is in the unexpected. SIMNET technology is still finding new applications in the military and civilian marketplace.
Early on, this team technology was oriented to training and readiness challenges--fight the present--because that was the prime concern of the Army leadership. Later it expanded into the acquisition business--fight the future--of the Department of Defense. The Pentagon decision makers were interested in more effective and lower cost ways of developing doctrine, material, tactics, and weapons systems, including analysis and testing. Thus, ARPA's initial plan included a developmental testbed that would be networked to industry and academia. This original 1985 diagram shows the vision for such a SIMNET testbed.
THE SIMNET TESTBED
Now, the Army has a network of battlefield development laboratories located at the proponent schools and the material development laboratories. This concept was demonstrated in October 1992 by the Chief of Staff of the Army for the four-star generals in the Army. The idea was to introduce the power of networked simulation in doing the business of preparing for war. Hopefully, the other services will join in to make a truly joint warfighting virtual environment.
See image above for THE BASIC NETWORK ARCHITECTURE
DON'T BE TRICKED BY THE NAME
Originally, in 1978, Thorpe used the term, "tactics development network/center" as a vehicle to introduce the future technology. Later in 1983, the name SIMNET became an acronym for simulator networking. In 1990, when ARPA transferred the program to STRICOM, they changed the name to Distributed Interactive Simulation (DIS). The 1992 Defense Science Board on Simulation, Readiness and Prototyping coined the term Virtual Simulation. Later General Paul F. Gorman framed the model of Live-Virtual-Constructive Simulation. And the current moniker, STOW, was fashioned by Colonel Robert Reddy. Today in commercial applications created by the entertainment industry, we find names like virtual Reality, virtual Environment, three-dimensional graphics, realistic simulations, etc...
Regardless of the name, the core SIMNET technology is still a large synthetic environment which can be entered in real time from stations world-wide. The technology allows many different ways to think about new approaches to doing business, military or civilian. Networks allow widely-dispersed participants to be brought together on a virtual piece of the Earth without anyone leaving home. This ubiquitous chart is a useful depiction of the flexibility and speed in visiting any spot on the globe including denied locations.
See image above for The grape chart
A THREAT TO INDUSTRY
SIMNET was a threat to the stand-alone simulator industry. In the spring of 1986 when the technology was being applied to helicopters, the ARPA program came under vicious threats by lobbyists from the large manufactories of flight and conduct-of-fire simulators. At the time, computer image generators, the system that paints the scenes of the environment, (e. g., the pilots view through the canopy or the vision block of the tank), were costing millions of dollars. The magic of SIMNET was the production of an affordable computer image generator that cost in the low hundreds of thousands of dollars, i. e., a ten-fold reduction in cost. The big defense contractors saw these low-cost CIG as a threat to their established market. With their well-placed lobbyist, they fought this up-start research effort. As a result, the ARPA program manager spent many unnecessary hours answering Congressional inquiries, General Accounting Office and Inspector General investigations. The industry kudos go to small business. Bolt, Beranek and Newman, Inc. and Perceptronics, Inc. with Delta Graphics, Inc. were the contractors that delivered the SIMNET development system. To be charitable, big industry was no help.
Now many of those same "critics" pose as self-appointed discoverers who want more than anything to say they saw the next wave coming before anyone else--or maybe atone for their earlier sins. The upshot is that SIMNET technology has changed the industry. At the 1985 Industry/Interservice Training Systems and Education Conference (I/ITSEC) only one SIMNET paper was permitted to be presented and one SIMNET display was allowed and then in an overflow hotel. Compare this to the 1994 I/ITSEC in which virtually every individual paper and company display had a reference to DIS. Meanwhile, during this short period of less that 10 years we have seen CIGs get better, faster, smaller, more reliable and cheaper. Today the market is loaded with high quality high-powered CIGS in the low hundred thousand dollar range. They continue to get better and cheaper, and competition is great.
COURAGEOUS DECISIONS
One may ask how such a high risk development could survive in this hostile environment? The short but straightforward answer is courage. Courageous decisions by giant leaders who listened to their own conscious and did the right thing for future warfighters. The leaders understood the critical need for a better technology for collective training and they had confidence that ARPA would deliver the same. As one looks back over the fast-paced history of SIMNET, there were clearly some moments when the leadership could have taken the easy route and killed the program. These wise leaders have never been given proper credit for their courageous decisions. Clearly, the SIMNET program was the recipient of decisions by giant leaders with strategic visions. This chart reflects the long-range vision of the SIMNET decision makers in what Dr. Elliott Jaques cites in, "Executive Leadership," as the requisite "time-span" for top-level leaders in managing third order categories of complexity.