Innovation in robot technology is expected to be one of the most effective solutions to social issues faced by Japan, such as the declining birthrate and aging society, and disaster response. Future vision in various fields is stated in the government documentation, "Japan's Robot Strategy --vision, strategy, action plan--" (2015). Shortly, robots will be required to perform a variety of tasks in the diverse, dynamic "open environment", rather than routine tasks in a pre-arranged "closed environment" such as factory production lines. However, there is a big gap between the high expectations for robots and the current capability of robots. While robots can reproduce predetermined tasks in a "closed environment" precisely and repeatedly, they have not been able to perform actions that are easy for humans to take, such as folding clothes, going up and down stairs by grabbing handrails, and grabbing daily necessities in an "open environment".

Conventional robots use sensors to precisely measure and model information in unknown spaces and act according to plans based on logical calculations. This is effective for routine tasks such as the ones on factory lines, but robots have many difficulties coping with the open environment in which we live and work. In an open environment, the behavior of the robot collapses due to the following factors:

• Sensors cannot acquire all the information
• Various elements are intricately related and difficult to model
• Situation changes from moment to moment

The traditional approach to these problems is to reduce the indefiniteness of the open environment owing to improvements in computational power and sensor accuracy. While it has expanded the range of robot applications, the ability to adapt to the open environment is essential for further expansion of the robot's activity. In addition, for robots to play active roles in human society, society must be ready to accept robots in turn. It is necessary to consider, from the early stages of research and development, how robots might affect society and the economy before being introduced into society.

With this background and a view to the future social implementation, we propose to promote "basic research on robotics that can adapt to an open environment" as well as "socio-economic impact assessment from an early stage of R&D". Basic research should be promoted from two approaches: "adaptive intelligence" and "adaptive body", by not promoting research on the intelligence and physical aspects of robots individually but promoting research on "robots as adaptive systems" by collaborating with both sides. Regarding the socio-economic impact assessments, four research themes should be examined: "sense of security with robots," "protection of privacy," "interaction with society," and "distribution of responsibility as a new legal system".

As for the basis for the development of robotics research, we propose the "construction of an open platform to accelerate basic research" and "competition-type R&D". The open platform for robot software and hardware will facilitate the entry of researchers from different fields into robotics research. Collaboration among researchers in different fields will lead to the establishment of a theoretical foundation for system integration. In addition, competition-based R&D will enable the reproduction and validation of research results, as a result, facilitating the bridge from basic research to applied research.

We proposed the research and development of robots that can flexibly adapt to open environments, as well as efforts to improve social acceptance and rulemaking for robots from the basic research stage. The early realization of the social implementation of such robots in various application fields will contribute to the enhancement of Japan's industrial competitiveness. At the same time, we strongly hope that this will be an effective solution to social issues facing Japan, such as the declining birthrate, aging population, and frequent natural disasters.