MODIFICATION
A -- DEFENSE SCIENCES RESEARCH AND TECHNOLOGY
- Notice Date
- 8/12/2002
- Notice Type
- Modification
- Contracting Office
- Other Defense Agencies, Defense Advanced Research Projects Agency, Contracts Management Office, 3701 North Fairfax Drive, Arlington, VA, 22203-1714
- ZIP Code
- 22203-1714
- Solicitation Number
- BAA01-42
- Response Due
- 12/15/2002
- Point of Contact
- Ralph Chatham, DARPA Program Manager, Phone (703) 696-7501, Fax (571) 218-4553,
- E-Mail Address
-
rchatham@darpa.mil
- Description
- BAA 01-42, ADDENDUM 10, SPECIAL FOCUS AREA: TRAINING SUPERIORITY. The Defense Advanced Research Projects Agency (DARPA) seeks proposals for innovative research and development of technologies to transform military training by providing continuously-available, on-demand mission-level training for all forces at all echelons. We propose to do this by creating an environment where there will be a set of continuously available virtual training wars in which any unit or individual can participate via personal computer-based systems that will teach and exercise users in the cognitive aspects of their warfare areas. INTRODUCTION. In a 2001 report, Training Superiority and Training Surprise, the Defense Science Board described a U.S. training superiority, developed over the last 35 years, that consists of creating a large-scale mock war within an instrumented range. Pilots or Army units ?fight? against a dedicated opposing force (OPFOR), attached to the center, who create a better-than-real enemy. After each engagement the units themselves and the OPFOR generate critical feedback in no-holds-barred after action reviews (AARs). Immediately after a unit completes about two weeks of combat training center (CTC) engagement simulations, it is among the best-trained units in the world. A few months later, however, skill decay and personnel turbulence has reduced the unit competence dramatically. Moreover, this kind of training is available to only some of our forces every few years. In short, there aren?t enough real or training wars to prepare all our forces in all the needed skills all of the time. In this program, we hope to bottle the cognitive learning parts of the CTC experience and export it electronically to all our active and reserve/Guard military forces, and eventually to civilian responders. If we succeed, our warriors, our staffs, and joint forces will want to and will train continuously in and for all kinds of conflicts, including in areas not well addressed by current training, e.g. mine, electronic, and no-plan come-as-you-are warfare. A POSSIBLE VISION OF FUTURE TRAINING. Imagine a future where a strike pilot deployed to a remote airbase or on an aircraft carrier, instead of playing a future computer game, checks into the DARPA-Legacy Continuously Available Training War (DARWARS) with a message ?I want to train.? His computer recognizes him and notes that he has not read the AAR from the previous session. While it enforces the reading (perhaps by monitoring the direction of pilot?s gaze) the system looks locally and globally for a mission to train that pilot on the key warfare elements he needs. The pilot and/or the DARWARS system collects other live participants and fills the gaps with computer-generated avatars. The pilot and his local team may interact with logistic, air control, service and joint staffs, and ground forces. An Army forward observer at Fort Hood, who, on his own personal computer, is practicing his tactical Pashto language skills with an avatar of a local ally, might direct the strike aircraft where to drop their bombs. The DARWARS system might detect, from user records, a need for specific soldier training. For example, it might automatically alter the DARWARS scenario to encourage a pilot to make a mistake that gets him shot down, thus providing a now-virtual injured pilot as a subject upon whom the forward observer can refresh his point-of-wound-care skills. If the pilot stays where he belongs, the system might shoot the avatar of the foreign ally to provide a suitable wound for the Army forward observer to treat. As each user checks out of DARWARS to perform duties in the real world (or get some sleep), the system, in collaboration with the user, would automatically generate an after action review to be used by the trained users, the warfare area specialists, and the DARWARS system to improve future training. In addition, the users themselves could write and contribute software to create new training scenarios, or warfare lessons to improve the local or global DARWARS training structure. THE DARPA TRAINING SUPERIORITY (DARWARS) PROGRAM. For an annotated viewgraph presentation of the program vision see the DARPA DSO website at http://www.darpa.mil/dso/ under ?Future Areas of Interest.? The goal of the initial three-year project is to start a paradigm shift in military training by providing a new kind of cognitive training experience for units and individuals based on continuously available wars. To reach this goal, we will start with a set of local, smaller-scale versions of the future architecture suggested by the vision paragraphs above. We will then join separate training systems to demonstrate scalability. The program will embrace two key areas: 1. Create scalable Last-Meter Training Systems (LMTS). These recognize that it is not enough to deliver training data to a local computer via a network or compact disk; the last meter between the computer and the human must be bridged to create understanding within an ?official intelligence,? the user. LMTS will train judgment and cognitive performance under stress and provide a 2-sigma improvement (over classroom training) in student, user, and unit performance. As the program progresses, we hope to combine multiple DARPA LMTS to scale up to massive, multi-user engagements. LMTS will measure, model, and react to student states. They might employ, as appropriate to the training task, any of the following: self-generating, non-intrusive user tracking and certification; auto-tutors, games and avatars to provide coaching, training, and performance aiding; ?intelligent? opposing and allied forces as well as neutral (potentially friendly or hostile) participants; self-generating and self-updating of trainable knowledge and content extracted (semi) automatically from user experiences and practice; automated scenario generation tailored to user (individual or unit) training needs; the ability to constitute (in real time) teams and task forces using other players and units encountered in the engagement; and a simulation infrastructure using reusable objects, agents, and standards. LMTS will train meta-cognitive and learning-to-learn skills. Development of LMTS must be accompanied by the creation of a training toolbox that will allow the training establishment to slash both development costs and the time required to build new training-ware. Six focus areas have been selected for the initial set of LMTS. LMTS development proposals in other areas will be considered, but strong justification for the value to DoD of the area, user interest and the potential for contributions from the users must be provided. Work on an LMT should generate training development tools, advance the DARWARS architecture, and create a useful training system of high military utility. See the website for more detail on each: a. Deployable air strike mission trainer; b. Joint Task Force staff warrior; c. Rapid tactical language training for all; d. Combat medical training; e. Dynamic battlefield surveillance planning; f. Information Technology Management aboard ship or at a forward base. LMTs are not envisioned as high-end, high fidelity trainers. They are to use readily available computing and display hardware. Adjunct hardware, such as student/user state sensors must, by the end of several years of development, be low cost and easy to use. 2. Create a DARWARS Architecture to populate with, and link stand-alone LMT systems. This will combine real-time generation, maintenance, updating, and modification of the training war scenario. The architecture should consider the learning and social implications of massive multi-player (MMP) games including their compelling nature, team-forming and team support, and encouragement of contributions by the users to local and global software. The architecture must be scalable to allow simultaneous training of teams and individuals on all levels (from E-1 to Flag officers). It must monitor skill progression. Training value will, however, be the driving factor in any DARWARS architecture. Ideas of general value to all LMT areas and cross-cutting technology developments that will enable the DARWARS architecture are encouraged with or without teaming. Cross-cutting technology development might include tutor authoring languages, tutorial dialogue management, automated scenario generators, continuous student-state monitoring capabilities, real-time assembly of simulation objects into meaningful presentations, techniques that identify ace performers early-on, development and assessment of the psychometric properties of simulations, realistic human behavior representation, etc. Work in these areas may be proposed as separate tasks; however, development of these technologies within the context of one or more LMTS is preferred. Additional areas of interest for development within either an LMTS and/or the DARWARS architecture include the following: A) Systems that automatically create in real-time and re-use scenarios so that DARWARS can encompass multiple wars, simultaneously track and ensure achievement of local and global training goals, allow rapid (often automated) insertion of new scenarios and technologies, and automatically place the student, users, and units at the right point in the ?training space.? B) A capability to generate and present automatically (semi-automatically at first) AARs that will provide assessment and feedback to/on all of the following: the learner/user, the training system, and warfare lessons learned. AAR generation should provide metrics for distinguishing between sound warfare actions and decisions and those that arise from luck or from DARWARS artifacts. HOW TO PARTICIPATE IN DARWARS. We encourage submission of white papers that encompass one or more of these program areas. White papers may include any single areas or any combination of the above areas. We encourage the formation of interdisciplinary teams to develop solutions to the problems posed by the development of DARWARS. We may ?mix and match? from among the white paper submissions in order to select the best combination of ideas. Prepare the initial white papers in the following format: I. Cover Sheet (include significant proposer information: Title, organization(s), PI, mailing address, email address, fax and phone numbers). II. Technical submission, (ideas, technical approach to achieve goals, 12-18 month milestones, 3-year tentative milestones; Be specific; eschew boilerplate text) 2-4 pages. III. Company resume, 1-page. IV. Cost estimate, 1-page. Open architecture and knowledge sharing is critical to the success of this program. Intellectual property developed under this program will be shared with other participants under a shared intellectual property agreement. Intellectual property created and owned by a proposer before joining the program will be protected. White papers received within 30 days of the publication of this Addendum will be evaluated immediately upon receipt. After review we will request revised white papers or full proposals. White papers received after the 30-day time period will be considered on a case-by-case basis. Notwithstanding the disposition of white papers, DARPA will respond to full proposals expeditiously. Regardless of the recommendation of the white papers, the decision to submit a full proposal is the responsibility of the proposer. The deadline for proposals is 12 December 2002. We encourage communications with the Training Superiority/DARWARS Program Manager, Dr. Ralph Chatham, rchatham@darpa.mil, 703-696-7501, during the preparation of white papers and full proposals. Submit white papers as email attachments sent to dsobaa01-42@darpa.mil. Word 97 or higher is recommended (Embedded text and .pdf files are acceptable) or by U.S. mail: (DARPA DSO, Attn: Training Superiority -BAA01-42 -Addendum 10, 3701 N. Fairfax Drive, Arlington, VA 22203). Facsimile transmissions will not be accepted. Include the phrase ?Training Superiority? in the title. Please refer to the original BAA01-42, dated September 4, 2001, for additional information.
- Record
- SN00138914-W 20020814/020812213844 (fbodaily.com)
- Source
-
FedBizOpps.gov Link to This Notice
(may not be valid after Archive Date)
| FSG Index | This Issue's Index | Today's FBO Daily Index Page |