The i-Walker: An Intelligent Pedestrian Mobility Aid - Computational Intelligence in Healthcare: Advanced Methodologies - page 101

communication, mobility, self-care, interpersonal interactions and relationships as well
as community and social life. All these domains can be affected by aging or disabilities
and are the base of personal autonomy and the satisfactory participation on them re-
flects on the self well-being. Assistive Technologies (AT) are of special interest, as the
average age of the population increases fast [4,22]. AT can participate in these activities
in order to enhance the user's autonomy, gathering all the environmental information
and making use of it properly.
Our idea is based on the notion of a Shared Autonomy between the user and its own
agent-based mediator with any information system at hand. Existing telematic health-
care systems that provide integrated services to users are not, to our taste, enough flex-
ible to allow a real personalization and maybe now it is too expensive to change them.
The shared autonomy concept is scarcely explored in literature and often it is mis-
understood as shared control ( e.g. , [27,14]). In the personal autonomy and disability
context, two different scenarios of the shared autonomy can be elicitaded.
- People presenting mainly physical impairments are able to define their own goals,
but due to their restrictions they usually are not able to execute them, suffering a
limitation in their autonomy. In this scenario the contribution of AT focus on phys-
ical devices, mostly mobility hardware, that allow them to reach their objectives.
These devices may be controlled by multi-agent systems or through an agent su-
pervised shared control if the user motor capabilities are not severely damaged. In
this scenario, user interfaces are very important to detect the user intention, which
is critical to define goals for the wheelchair to be able to assist him/her.
- People presenting mostly cognitive impairments may require a different kind of
assistive aids, which may lead even a more relevant role in the sharing of personal
autonomy. In this scenario the user probably does not have very clear goals or is
not capable of achieving them because he/she cannot remember how to do them.
In these cases, AT may empower and complement their autonomy using agents
that offer them a set of services, like reminding what kind of activities they can or
should perform at a certain moment of the day or pointing them out how to achieve
these activities. The main idea is to offer the users a set of cognitive aids, either
rational or memory based, that can ease their daily living.
Roboticists have developed a number of mobility-enhancing assistive technologies.
Most of these are active aids, meaning as said before, that they share control over motion
with the user. Most are aimed at obstacle avoidance and path navigation [26][10][15].
4-Walkr
With this context in mind, we introduced in [6] the design of an integrated architec-
ture aimed at helping citizens with disabilities to improve their autonomy in structured,
dynamic environments. The main element of this architecture is an intelligent agent
layer that mediates between different technology components (robotic devices -as the
i-Walker - ubiquitous computing, and interfaces) in order to provide the subject with the
necessary degree of independent mobility to benefit from different assistive services and
to reach goals determined by either the subject himself/herself or by medical staff.