Autonomic Robotics for Smart Environments
Introduction
The design and development of smart home technology and algorithms to support people with reduced physical or cognitive ability has received much attention in the past few years. In reported studies, monitoring a person in a home environment is achieved with the use of embedded sensors (Atallah, 09). A smaller number of studies have reported assistive robots developed to complement caregivers e.g. Nursebot (Pollack 02). Little has been done in the way of a mobile robot cooperating with the intelligent environment to provide assistance (Mastrogiovanni 09, Deegan 07). Assuming a future where the Smart Environment is responsible for monitoring an individual's well-being, yet parts of the systems will fail. Examples in the press include demensia patents wandering out a back door where the sensor has failed and no alarm raised. What is required is a combined approach, that not only does the system monitor the environment but also itself (Sterritt et.al. 10).
Technology in the form of embedded sensing, is gaining acceptance as a means to assist persons with reduced physical or cognitive capacity in their day-to-day living. Robots, and more specifically mobile robot assistants, have yet to have gained acceptance in Europe but with the predicted increase in the retired population robotic assistants can help alleviate the problem of shortage of carers. A mobile robot working within a smart home must be capable of navigating, recognising typical objects found in the home. To be effective in a home environment, the robot must be capable of adapting and learning new skills or behaviours. Since reliance on the continued operation of such an environment will become critical, the robot may also act as part of the autonomic/self-management control loop (Sterritt etal 09). A challenge for robotics research is to develop adaptable robots that through learning can operate reliably in a dynamic environment.
Brief Methodology The aim is to monitor the inhabitant of a smart home with a robot and provide assistance when necessary; for example by locating a misplaced object. Likewise, in terms of providing Autonomicity (self-managing system), if part of the sensor system is not operating correctly the robot can be directed to investigate and potentially deposit temporary sensor to re-establish a full operating smart environment (SE). For the purpose of monitoring, the robot’s sensor complement will be augmented to include standard SE sensors; essentially making the robot a mobile SE sensor. Rather than pre-programming the robot, it must learn the skills. The proposed work will use autonomic computing techniques to monitor and effect the environment and the system itself.
Brief Work Plan Literature review in the areas of behavioral robotics, service robots and autonomic computing.
• Investigate requirements to support a person with mild physical and/or cognitive decline in the home and define a use case with input from stakeholders
• Develop a design framework to provide a solution for the use case
• Validation and assessment of a demonstrator with a demonstrator within the Living Lab
Context This project will be linked in with existing work and projects in the area.
Resources: The project will have access to several robots: Sputnik3, Pioneer 3DX, and a Nao
References
- L. Atallah and G-Z., The use of Pervasive Sensing for Behaviour Profiling - A Survey, Yang in Pervasive and Mobile Computing, 2009
- M. E. Pollack, et al: A Mobile Robotic Assistant for the Elderly, AAAI Workshop on Automation as Eldercare, Aug., 2002
- F. Mastrogiovanni, A. et al. Context Assessment Strategies for Ubiquitous Robots. In Proceedings of the IEEE International Conference on Robotics and Automation (ICRA 09), Kobe, Japan, May 12-17, 2009.
- Deegan, Patrick, et al. Mobile manipulators for assisted living in residential settings 2007
- Sterritt, Roy and Nugent, Chris (2009) Autonomic Computing and Ambient Assisted Living. At: AA-SES-IV: 4th IEEE International Workshop on Autonomic and Autonomous Space Exploration Systems (at SMC-IT), Pasadena, CA, USA. IEEE Computer Society
- Sang Hyoung Lee, Il Hong Suh: Bayesian network-based behavior control for skilligent robots. ICRA 2009: 2910-2916
- Sterritt, Roy and Bustard, David W. (2010) Autonomic Systems. In: Encyclopedia of Software Engineering. Taylor & Francis, pp. 111-117. ISBN 1-4200-5977-7
Personnel
Roy Sterritt, George Wilkie
Personnel Involved
First Supervisor: Sterritt, R Mr
Second Supervisor: Wilkie, FG Dr
Collaboration: This project does not involve collaboration with another establishment
Synopsis:
The project will involve undertaking a comprehensive study of incorporating autonomic (self-managing) and adaptive behaviour for Robots to ensure reliability of Smart Environments.