| We developed and evaluated a semi-autonomous mobile healthcare service robot named EL-E. I conducted four human studies involving patients with ALS with the following objectives: (1) to investigate and better understand the practical, everyday needs and limitations of people with severe motor impairments; (2) to translate these needs into pragmatic tasks or goals to be achieved through an assistive robot and reflect these needs and limitations into the robot's design; (3) to develop practical, usable, and effective interaction mechanisms by which the impaired users can control the robot; and (4) to evaluate the performance of the robot and improve its usability.;In the initial user needs assessment, I identified the needs and wants of the target users (recruited from the ALS Center of the Emory Clinic) through user interviews and field studies involving the documentation of occurrences of object retrieval difficulty. The results help to inform the design and development of the robot system, with respect to common objects to be retrieved, preferable methods of object return/delivery, and the acceptance of potential control interface, as well as to direct the planned evaluation activities to test the assistive robot's abilities to meet the users' needs. This needs assessment also led to the development and validation of a list of common objects ranked according to their relative importance in terms of user needs for activities of everyday living and, thus, potential targets for robotic-assisted retrieval.;To this end, an initial series of human evaluation studies were conducted to study two pragmatic issues---the design of a usable user control interface for users with varying motor limitations and the object delivery method of handing off a retrieved object to the human user. In the first study, three distinct user control interfaces were developed: (1) a modified, hand-held laser pointer; (2) a modified, ear-mounted laser pointer; and (3) a touch screen, graphic user interface on a portabe computing device. Users were asked to utilize these control interfaces to direct the robot to the three-dimensional coordinates of a target object for the purposes of object retrieval. The empirical results illustrated a high success rate of object retrieval (94.8%) and a very high level of user satisfaction. Perhaps not surprisingly, user preferences for the control interface were highly correlated with users' quantitatively measured motor capabilities, with upper limb mobility an important determinant of interface preference. More importantly, however, three viable control interfaces were developed to accommodate the variability in ALS patients' motor capabilities. Additionally, the study illustrated that the general paradigm of conveying 3D object location information to the assistive robot, regardless of the interface mechanism used, can successfully result in task completion without any required changes to the actual robot design and functionality.;The second user evaluation study examined two delivery mechanisms---direct delivery to a user's hand and indirect delivery to a nearby surface---to determine the limitations with robot-assisted object delivery. Overall, the robot successfully delivered objects with a success rate of 98% for indirect delivery and 78% for direct delivery (with an overall success rate of 88% or 126 out of 144 trials across all conditions). The results indicate that indirect object delivery, the preferred method for some users, provides a robust and reliable (albeit slightly inefficient) object delivery method. (Abstract shortened by UMI.)... |
