BAA-09-01-PKS-Amd30.doc

Broad Agency Announcement (BAA)

NR: 09-01-PKS, Amendment 30 Sensor Technology Research, Development, Test & Evaluation Open-Ended Broad Agency Announcement (STROEB) II BROAD AGENCY ANNOUNCEMENT TITLE: Sensor Technology Research, Development, Test & Evaluation Open-Ended Broad Agency Announcement (STROEB) II BROAD AGENCY ANNOUNCEMENT TYPE: This is Amendment 30 to the Open-Ended Broad Agency Announcement.

BROAD AGENCY ANNOUNCEMENT NUMBER: STROEB II BAA 09-01-PKS Amendment 30 AMENDMENT TOPIC TITLE: Compact ATR & Sustainable Environment (CASE) TECHNICAL POC: Pat Marshall, AFRL/RYAR, Phone (937) 938-4294, email: pat.marshall@wpafb.af.mil. Mailing address: 2241 Avionics Circle, Wright-Patterson AFB, OH 45433 CONTRACTING POC: Jennifer Kleiv (Primary), Contract Negotiator, (937) 656-9828, email: Jennifer.Kleiv@wpafb.af.mil; Paul E. Smith, (Secondary), Contracting Officer, (937) 255-4279, Email: Paul.Smith@wpafb.af.mil. Mailing address: AFRL/RQKSR, Bldg 167, Area B, 2310 8th St., Wright-Patterson AFB, OH 45433-7801

Background: Identifying military combat targets from sensor (e.g., EO, IR, SAR, etc.) data is a very difficult task due to vast number of variables influenced by natural and man-made environmental factors. This challenge is amplified for moving objects in target-like clutter usually seen in urban scenes with temporal occlusions. Legacy exploitation algorithms tackle this problem by requiring extensive training data sets of all objects in the scene as representative sets of classes and sub-classes that are measured and/or simulated under a vast number of operating conditions. However, this creates a situation inherently complex to maintain and, as a result, very costly. Another non-technical issue deals with proprietary rights. Most platforms currently in DoD inventory are comprised of contractor proprietary sub-systems where the government program manager is constrained to return to the development contractor for all enhancements . This slows the pace of innovation while at the same time driving up yearly operational costs. All of these issues, when combined, prohibit maintaining current target set systems in a timely and cost-effective manner for operational platforms.

Objective and Scope: The primary objective of this effort is to develop sustainable multi-phenomenology ATR based on low fidelity, compact, and inexpensive database technology. Data volume, cost, and processing timelines must all be reduced by comparing sensed data to easily derived features from a minimum-parameter database, which can also be quickly and easily extended to accommodate new targets. As such, a new construct is required whose priority is to take into account the size and complexity of the database at the earliest stages of automated target identification algorithm development. These new ATR technologies will utilize compact, simpler, yet robust databases as inputs to totally new signature exploitation algorithms across various sensor modalities. This includes using saliency technology as an enabler of compact, reduced-cost, sustainable object databases on operational data, and integrating with realistic exploitation algorithms. This is a big departure from other efforts and recent developments which require very large data collections for target model validation and target library training. It is envisioned that feature-based classifiers that uses a small set to represent each target in the library will be developed by this effort. An example for ground vehicles may be physical features of the target such as: length, width, height, cab position, wheel count, axle spacing, etc. Simultaneously, very fast/low-maintenance target classification algorithms will be developed that utilize the new compact data bases. Another example for large EO images may potentially be to use much simpler 2-D target-size metrics, allowing correlation algorithms orders of magnitude more efficient in speed-up time as compared to processing full image frames. This could allow operators and analysts to quickly update databases with new targets of interest using only a small number of critical characteristics (e.g. height, length, width). The end result eliminates a major problem with current technology which requires a long “logistics tail” which is inherently difficult, time consuming, and costly to add new targets. Tradeoffs must be identified and addressed. For example, loss of fidelity may be overcome simply by flying the sensor at a lower altitude to increase target resolution. Sensor operating constraints (SOC) are important throughout this process. Exit criteria to this effort will be a function of financial cost and amount of time it takes to create, update, and maintain these new target libraries. Lastly, innovative solutions to the proprietary problem will also be investigated. Contractor is expected to be an expert in all areas from research in a laboratory environment to flight testing prototypes on manned and unmanned air platforms.

Programmatic: A Cost-Reimbursement/Fixed Fee contract type is contemplated with multiple awards to be split among $24,350,000 total for this seven-year program. The anticipated amount is an estimate only and is NOT a promise of funding. Funding is uncertain and is subject to change. Changes in availability occur at the discretion of the Government.

The two-step process will be used for this solicitation. As such only white papers will be requested from the upcoming SOO publication. White papers will then be evaluated and only those selected may be asked to submit a formal cost and technical proposal approximately thirty (30) days after white papers are received and evaluated.

DO NOT SUBMIT ANY DOCUMENTAION AT THIS TIME.

THIS IS NOT A CALL FOR WHITEPAPERS NOR PROPOSALS.

ANYTHING RECEIVED WILL BE DISREGARDED.

� DoD Defense Science Board, Task Force Report: “The Role of Autonomy in DoD Systems”, July 2012.