DARPA wants creative concepts to solve complex catastrophes

The Defense Advanced Research Projects Agency is seeking innovative ideas and concepts that can advance the area of strategic collaboration, according to a request for information issued by DARPA.

The agency wants interested individuals and organizations to submit white papers on technologies and concepts that could form the basis for a DARPA program to develop a network-enabled collaborative environment. Ultimately, the networked environment would bring together large numbers of people with different experiences, cultures and expertise to address the complex problems of large-scale disaster recovery and relief operations.

The recent and increasing prominence of Stability, Security, Transition and Reconstruction (SSTR) and Humanitarian Assistance and Disaster Relief (HADR) operations pose new challenges for the U.S. military, according to DARPA. These operations involve a large, diverse mix of military organizations, nonmilitary government organizations, regional and international government agencies, nongovernmental organizations, private volunteer organizations, individual volunteers and the local population, the RFI states.

These disaster relief and reconstruction operations exceed the ability of any one actor or organization to solve or even comprehend, DARPA officials said. Participants in SSTR/HADR operations have to collaborate across domains, organizations, cultures and languages.

As a result, the DARPA RFI seeks information in the following areas:

• Semantic glue: This capability will provide individuals from diverse organizations and cultures, speaking different languages and having different goals, the ability to collaborate in support of SSTR/HADR operations. Semantic glue refers to the collection of technologies and capabilities that will enable strategic collaboration by providing adaptive and interactive representation schemes ? such as language, icons and geospatial graphics ? for exchanging information across a wide array of devices ? ranging from cell phones to desktop computers ? in environments with varying network bandwidth and reliability, according to the RFI.
  
• Ad hoc, dynamic networking on diverse, unstable networks: Ideally, SSTR/HADR responders should leverage existing networks and devices to support strategic collaboration. This requires the ability to perform ad hoc, on-demand networking on public and private networks, which may be unstable and evolving. Participants need to be able to detect the health and status of the networks and manage the flow of information. Individuals and organizations must be able to register their existing devices and platforms, which may not be interoperable. This must be accomplished with minimal technical staff and often by nontechnical operators.
  
• Mobile computing applications to support local optimization: Recent Web-based examples have demonstrated that people can dynamically network to achieve such common goals as eBay-style online auctions, social networking and online volunteer activities. An analogous set of tools and capabilities could be developed to support SSTR/HADR operations, allowing both formal organizations and individuals to locally match their resources and needs.
  
• Understanding (human) network performance: The focus here is to provide a higher-level understanding of emerging human networks, their capabilities and capacities that can be understood by average people on the scene charged with making key resource decisions.
  
• System level issues: A system that has all of the necessary capabilities to meet SSTR/HADR challenges has to be able to operate in both reach-back and stand-alone modes. It needs to be robust enough to withstand connectivity disruptions, and maintain system and information integrity and security. It also needs to be persistent across many operations to enable learning. For example, last year's tsunami becomes this year's game/simulation, which becomes next year's deployable system.
  
• Mensuration: The resulting system would be instrumented to support continuous adaptation, feedback and optimization of resources. Embedded real-time metrics could support machine learning for both human networks and the communication networks upon which they depend.