The CCC-sponsored initiative in robotics, led by Henrik Christensen, has made great progress and provided a model example of a CCC initiative.  Having finished their series of workshops and developed a roadmap, they are now bringing targeted portions of that roadmap to NSF, NIST, DARPA, NIH and the Office of Science and Technology Policy.  They are also organizing a U.S. Congressional caucus on robotics to take place in March.  Additionally several companies have expressed an interest in engaging in a broader effort on robotics across United States.

Back in early 2008, they began organizing four workshops, one each in four topical areas of robotics: manufacturing and logistics, healthcare and medical robotics, service robotics and emerging technologies.  More than 100 people attended these workshops, representing a mix of industry and academia.  Preliminary workshop reports were made available to the community, and an online discussion board provided a forum for further input.

A sampling of the four workshops’ findings:

• In manufacturing it is evident that the main applications so far have been large-scale production of entities such as cell phones and cars whereas small-scale production has received limited attention.  It is further evident that processes such as logistics and material handling have significant potential for use of robotics, but so far little attention has been devoted to such applications.  There is a need to consider new methods for easy programming of robots, and further integration of sensory information to enable robust and safe operations.  Less than 5% of all industrial robots today use sensors as part of the primary control system.
• Medical robots are today widely used for prostate surgery and are also gaining momentum for cardiac procedures and hip replacement.  The main motivations are faster recovery, improved quality and a reduced risk of any side effects.  The potential for medical robotics is very significant.  Related to healthcare there is also the use of robots for rehabilitation as it enables a higher degree of customization to individual patients and faster initiation of training.  In addition the engagement with robots is sometimes easier than interaction with humans due to privacy and scheduling considerations.  Wider adoption of healthcare robotics calls for new methods in machine learning, human robot interaction and flexible mechanisms for physical interaction with humans.
• Service robotics has two aspects: professional and domestic.  Professional robotics involves for example agriculture, forestry, mining and harbor automation. The number of people involves with agriculture and related industries is decreasing while the demand is increasing and there is a need to further automate the industry to remain competitive.  For domestic services there is a need to provide cleaning, surveillance, life sign monitoring, remote video, etc to assist people in their busy lives, but also to provide key functionality to enable people to remain in their homes as mobility and mental capabilities are reduced with age.  In service robotics it is characteristic that users have no or very limited training and the systems must be intuitive / easy to use.  In addition there is a need for flexible integration with existing technology (a scalable integration strategy). Finally there is a need for navigation and flexible perception to allow deployment in natural environments (e.g. homes).
• In emerging technologies that are several opportunities as sensing become ubiquitous, more flexible mechanisms are designed and new technologies such as nano become available.  The access to complex computing with a limited footprint allows deployment of AI in new settings.  The use of machine learning and new types of interfaces with a high degree of connectivity opens entirely new opportunities for use of robotics. Not to mentioned new actuation methods.

Robotics in general is characterized by a significant economic potential and has research opportunities across the entire spectrum from basic to applied.  There are clear short-term opportunities in areas such as medicine and manufacturing and at the same time there is a potential to create an entirely new industry for cognitively endowed robots with richer interaction with the world.

The workshop reports are nearly in their final form.  You can watch for them and other updates at http://www.us-robotics.us.

–Andrew McCallum