N101-076 TITLE: Platform for Developing and Evaluating Spatio-temporal Cognition in Autonomous Agents

TECHNOLOGY AREAS: Information Systems, Human Systems

OBJECTIVE: Develop and demonstrate a platform for building and evaluating spatio-temporal reasoning capabilities of virtual autonomous agents.

DESCRIPTION: Land-based and airborne semi-autonomous systems are being increasingly used in critical missions such as search and rescue, information surveillance and reconnaissance (ISR), and explosive ordnance disposal (EOD). However, interactive navigation capabilities for such systems are cumbersome to use and can limit the scale of operations.

The existing methods of human interaction with robots and other autonomous systems such as unmanned aerial vehicles (UAVs) are platform specific and severely limited (Burke et al, 2004). For instance, Lauria et al (2002) illustrate the development of a natural language interface that employs shallow text processing. Likewise, Skubic et al (2007) show the utility of sketch based interaction for controlling a team of robots. However, these approaches ignore the connection of human robot interaction to spatio-temporal cognition, and world knowledge as a basis for more robust and efficient interaction. Moreover, presently evaluating the impact of such interaction methodologies is difficult due to heterogeneity of approaches and make the comparison of alternative approaches infeasible (Olsen, 2006).

Navigational and extra-navigational autonomy of these systems needs to be increased along with a provision of robust interfaces for natural language and sketch-based interaction using deeper spatio-temporal reasoning abilities (e.g., Roy, 2005). Developing navigation capabilities and corresponding interaction techniques for specific platforms can lead to brittle and unportable spatio-temporal reasoning systems that fail to perform outside their designed environments. Furthermore, the development and evaluation process of such capabilities in physical systems can be time consuming, expensive, and can result in unpredictable performance. Therefore, an integrated platform for developing and evaluating interactive spatio-temporal reasoning capabilities of cognitive agents is needed. The desired capabilities for such a platform are as follows: (1) author and generate scenes and scenarios that represent the target environment and the navigational task for autonomous agents (2) embed or integrate autonomous agents in a plug-n-play manner (3) interact with autonomous agents; such as submit high-level navigational commands in one or more modes such as natural language and/or sketching (4) observe and record the reasoning and activities of autonomous agents in their environments (5) measure, analyze, and report the performance of virtual autonomous agents.

PHASE I: Survey the representation approaches for portable, interactive, spatio-temporal reasoning and select a candidate spatio-temporal representation and reasoning system. Develop component architecture for a system that enables development and evaluation of spatio-temporal reasoning in autonomous agents and demonstrate the feasibility of desired capabilities.

PHASE II: Develop a detailed design and implement a prototype of the system proposed in Phase I. Demonstrate the ability of the prototype to meet the desired capabilities by creating and executing several navigation problem scenarios of varying complexity.

PHASE III: Fully develop the system into one or more commercial products. Transition the technology for use in the (X: Navy FNC?) and for commercial applications.

PRIVATE SECTOR COMMERCIAL POTENTIAL/DUAL USE APPLICATION: Robust interactive multimodal interaction for navigational and extra-navigational capabilities in virtual and physical autonomous agents has wide-spread application both in other branches of DoD and industry. Potential applications include interactive training and tutorial systems, gaming, and simulation.

REFERENCES:
1. Burke J.L., Murphy, R.R., Rogers, E., Lumelsky, V.J., & Scholtz, J. (2004). Final report for the DARPA/NSF interdisciplinary study on human-robot interaction. IEEE Transactions on Systems, Man, and Cybernetics, Part C: Applications and Reviews, 34(2):103�112.

2. Lauria S., Bugmann G., Kyriacou, T. & Klien E. (2002). Mobile robot programming using natural language, Robotics and Autonomous Systems, 30, 171-181.

3. Olsen, D.R., & Goodrich M.A. (2006). Metrics for evaluating human-robot interactions, Proceedings of the 1st ACM SIGCHI/SIGART conference on Human-Robot interaction, Salt Lake City, Utah, USA pp. 33 -40.

4. Roy, D. (2005). Grounding words in perception and action: computational insights, Trends in Cognitive Science, Vol 9 (8)

5. Skubic, M., Anderson, D., Blisard, S., Perzanowski, D., Shultz, A. (2007). Using a hand-drawn sketch to control a team of robots, Autonomous Robots, 22 (4): 399-410.

KEYWORDS: Virtual environments, autonomous systems, spatio-temporal reasoning

** TOPIC AUTHOR (TPOC) **

DoD Notice:   Between November 12 and December 9, 2009, you may talk directly with the Topic Authors to ask technical questions about the topics. Their contact information is listed above. For reasons of competitive fairness, direct communication between proposers and topic authors is not allowed starting December 10, 2009, when DoD begins accepting proposals for this solicitation.

However, proposers may still submit written questions about solicitation topics through the DoD's SBIR/STTR Interactive Topic Information System (SITIS), in which the questioner and respondent remain anonymous and all questions and answers are posted electronically for general viewing until the solicitation closes. All proposers are advised to monitor SITIS (10.1 Q&A) during the solicitation period for questions and answers, and other significant information, relevant to the SBIR 10.1 topic under which they are proposing.

If you have general questions about DoD SBIR program, please contact the DoD SBIR Help Desk at (866) 724-7457 or email weblink.