Pineau’s paper discusses a mobile robotic assistant, developed to assist the elderly individuals with mild cognitive and physical impairments, as well as support nurses for their daily activities. The software modules that comprise this system include: An automated reminder system; a people tracking and detection system; and finally a high-level robot controller that performs planning under uncertainty by incorporating knowledge from low-level modules and selecting appropriate courses of action.

The robot primarily interacts with the world through speech, visual displays, facial expressions and physical motion. One of the issues that this particular model has come across is that many elderly have difficulty understanding the robot’s synthesized speech as well as articulating an appropriate response in a computer-understandable way. This refers back to the same idea as in Eriksson’s “Hands-off Assistive Robotics for Post-Stroke Arm Rehabilitation”, where a happy pre-recorded voice and encouraging movements made the patient interested and inspired to perform the activity well. This is the same issue as social acceptance and therefore it makes sense that if the robots are going to replace certain human roles, than they should try to emulate a human in those roles as much as possible. If this issue is not addressed, than the elderly in general will not trust this type of technology.

The design of the system itself also needs to be flexible and adaptable, responding to the actions taken by the user. This is important, especially with the activities such as giving the patient their medication. This system needs to determine the most effective times to issue each required reminder, taking the user behavior into account, and any preferences explicitly provided by the user and the care-giver.

The final piece of the architecture addresses the problem of locating people by determining their x-y location relative to the robot. One of the algorithms used in this approach is with particle filters, which is addressed in Fox’s “Monte Carlo Localization: Efficient Position Estimation for Mobile Robots”. As the robot reduces its uncertainty, the number of modes in the robot pose posterior quickly becomes finite, and each such mode has a distinct set of people estimates. Another piece to finding a user would be the social interaction piece. In the experiment, is was demonstrated that the robot asking a clarification questions is in fact much cheaper than accidentally guiding a person to a wrong location, or guiding a person who does not want assistance.

This research does represent a feasibility of integrating assistive robots into the homes of the elderly and nursing homes. The patients will be accepting of the robots but some outstanding social interactions need to be addressed such as speech and general interaction with the user. It is more natural to communicate verbally then via a touch screen on the robot. After watching the video, this becomes evident in all the interactions.

Reference:

  1. Pineau,  Joelle. “Towards Robotic Assistants in Nursing Homes: Challenges and Results.” <http://www.cs.mcgill.ca/~jpineau/files/jpineau-ras03.pdf>