| Home | Publications | Proceedings |
Objectives: The ProVAR (Professional Vocational Assistant Robot) system is an assistive robot that enables individuals with a physical disability such as high-level tetraplegia to manipulate physical objects in a semi-structured office workstation environment. ProVAR allows its users to create and execute Activities of Daily Living (ADLs) and vocational support tasks that are complex yet robust. The goal of the project is to augment the user's autonomy and independence while simultaneously reducing the resources needed in terms of human attendant care in a vocational setting.
Methods: The robot arm, mounted overhead on a transverse, motorized track, can execute tasks created through commands given by the user, who manipulates a 3-D graphic representation of the workstation. The user previews the task on-screen in a live dynamic simulation of the robot calculated by the robot arm's controller before final execution of the task by the arm. The ProVAR interface combines a task creation and manipulation window with the 3-D world model. The representation of the user's work area contains a "live" robot model showing the position of the real robot and a "virtual" robot model that the user can move to set new goal positions. The user interacts with the ProVAR system through various adaptive access devices selected according to each user's specific physical capabilities and preferences. The ADL and vocational manipulation tasks that ProVAR can perform (medication, drinks, paper handling, disks, videotapes, etc.) allow the user to be without an attendant for a large portion of a working day, increasing overall independence and permitting a measure of personal privacy.
Results: User testing has aided in refining the interface and confirming the success of the design strategy. The test subjects were able to execute pre-programmed tasks and create new ones after two training sessions. Using the interface and creating new tasks were reported to be "very easy, very straight-forward." The test subjects gave enthusiastic responses when they were able to watch the robot arm perform a task in real life that they had created and tested on-screen in simulation.
Conclusions: The novel interface and controller designs are expected to lead to technology transfer of the ProVAR concept and an eventual assistive robot product. Additional user trials in a vocational setting are pending.
Acknowledgments: This work was supported by a Veterans Affairs Rehabilitation R&D Service grant, with additional support by the Stanford University Departments of Mechanical Engineering and Computer Science.