NSF and other agencies announce the National Robotics Initiative
The National Science Foundation (NSF) announced the National Robotics Initiative (http://www.nsf.gov/publications/pub_summ.jsp?org=ENG&ods_key=nsf11553) on 24 June; the solicitation is unusual in that it involves collaboration with the National Institutes of Health (NIH), the United States Department of Agriculture (USDA) and the National Aeronautics and Space Administration (NASA). Letters of Intent are required, and are due on 1 October for small proposals ($100K to $250K per year in direct costs for up to 5 years) and 15 December for "group large" proposals (from $250K to $1M per year in direct costs for up to 5 years, not to exceed $1.5M per year in total costs). The full proposal deadline is 3 November for smalls and 18 January 2012 for group larges. The solicitation is more complex than usual for NSF, in part because of the partnership with other agencies, and interested parties should read it carefully. I am bringing this solicitation to the Cyberling community because it provides an unusual funding opportunity for linguists and other cognitive scientists to collaborate with roboticists. The goal of the solicitation is "accelerate the development and use of robots in the United States that work beside, or cooperatively with, people"; such robots are referred to in the solicitation as "co-robots". It lists eleven "aims", two of which are:
• Pursue fundamental research in robotics science and technology and in supporting specialties in machine cognition, language understanding and production, human-robot interaction, perception, systems and other disciplines relevant to co-robot capability and performance.
• Explore how co-robotics designs can be enhanced by leveraging and integrating our understanding of human cognition, perception, action control, linguistics, and developmental science.
It also lists six "fundamental research topics", three of which are:
• Problem solving architectures that integrate reasoning, motor, perceptual, and language capabilities and that can learn from experience.
• Hybrid architectures that integrate or combine different methods, such as deductive, probabilistic, analogical, case-based, symbolic, or sub-symbolic reasoning.
• Computational models of human cognition, perception, and communication for commonsense or specialized domains and tasks, including acquisition and representation of contextual knowledge.