VIRTUAL PROTOTYPING OF REHABILITATION AIDS

Matthew T. Beitler, William S. Harwin, Richard M. Mahoney

Applied Science and Engineering Laboratories (ASEL)

Universityof Delaware/Alfred I. duPont Institute

Wilmington, Delaware, USA

ABSTRACT

This paper discusses methods of using virtual prototyping to assist with the development and customization of rehabilitation products. The difficulty of designing equipment for individuals with physical disabilities is that each individual presents a unique neuro-physiological picture. This research explores ways in which virtual device design and evaluation can be combined with new techniques in rapid prototyping to create efficient rehabilitation devices. introduction Due to the considerable diversity of users' physical abilities, there exists a need for designing devices which are user specific. This design cycle is complicated by the complexity of the device and the small market size. There are also numerous biomechanical factors that need to be considered. The techniques of virtual design and rapid prototyping are ideally suited for assessing these factors as well as rapidly designing and manufacturing one-of-a-kind rehabilitation devices. A method of designing and fabricating rehabilitation devices has been proposed which attempts to separate the user interface design from the supporting features design [5]. The fundamental assumption of this method is that most of these supporting features can be acquired from conventional design techniques, whereas the user interfaces must be customized for each individual. Wheelchair seating and controls, prosthetic arms and legs, as well as the Magpie [2] are instances in which this method has proven to be useful. Consequently we are attempting to develop a representation for design abstraction and associated reconfiguration mechanisms that may be used for the development of other configurable and extensible devices. The development of this representation involves the use of graphical computer models and interface devices to support the development of virtual mechanisms as well as facilitating the analysis of user tasks and user abilities. Background Assistive devices need not be complex for them to be useful in tasks of daily living. Simple designs are often more reliable, have a lower cost, and are more readily accepted. The Winsford Feeder [4] and the Magpie [2] are examples of widely used rehabilitation devices which have a simple design and are suitable for a wide range of disabilities. However, the major disadvantages of these devices are their inability to be readily adapted, their long manufacturing time and their high cost.

Figure 1: Rapid Prototyping of Rehabilitation Aids Development Cycle

Effective design of rehabilitation devices necessitates the quantitative assessment and feedback of the disabled person's performance as well as a method for quickly designing and prototyping assistive devices based on performance goals. Our complete approach (shown in Figure 1) to design and prototype passive, assistive, mechanical devices involves:

· the quantitative assessment of the form (geometry) and performance (kinematics, dynamics) of human limbs using sophisticated methods from computer vision and biomechanics · design of the assistive device

· evaluation of the device using simulation and virtual prototyping

· feedback from the consumer and associated personnel such as the therapist or physician

· actual prototyping of the assistive device

· evaluation of the function and performance of the device

· redesign based on performance

This paper focuses on the simulation and virtual prototyping aspects of the development cycle, which involves assessing the user's capabilities, rapid creation of virtual devices based on kinematic specifications and anthropometric data of an individual user as well as allowing the user to test and qualitatively access the capabilities of a prototype.

Figure 2: Virtual Prototyping Conceptual Diagram Method

The research into the virtual prototyping aspects of this project is being conducted using a Silicon Graphics workstation with a Spaceball 3D input device. To display graphical objects we are using JACK¿, a software application which displays and manipulates articulated geometric figures and includes a human body model which can be customized to a specific set of anthropometric variables. These components allow us to interactively control 3D graphical objects with simultaneous six degrees of freedom. A Spaceball¿ 6DOF isometric joystick, which has proven effective in controlling rehabilitation devices in the past [6], allows the designer and user to experiment with controlling the virtual prototype device. The different degrees of freedom of the virtual prototype device can also be controlled with keyboard inputs. Additional input devices are also going to be integrated into the system, including both a PER-Force Handcontroller and a Phantom¿ which will allow the system to give haptic feedback to the user.

Figure 3: Denavit-Hartenberg Parameters for the PUMA Robot

The first step involved in the virtual design process (shown in Figure 2) is the definition of the prototype device. The designer accomplishes this by defining the Denavit-Hartenberg parameters [5] of the device they are prototyping. The Denavit-Hartenberg notation is a common kinematic protocol used by designers for defining a device's movements. After the designer has inputted the Denavit-Hartenberg parameters (shown in Figure 3) our software automatically generates a 3-D graphical model of the device (shown in Figure 4) that can be manipulated by the user. This allows the designer to define which external inputs control the movements of the different joints of the device. The model of the device can be altered interactively by the designer, so that she can customize it according to feedback from the user and the evaluation process.

Figure 4: Virtual Model Generate from the Denavit-Hartenberg Parameters of the PUMA Robot

The last component of the virtual design process allows the designer to take the virtual model of the prototype and translate it into a format that can be inputted into a CAD design package (ProEngineer¿) to generate the physical prototype [3]. conclusion This research is part of a larger consumer focused design program that is investigating methods of product design that are adaptable to consumer needs, are cost effective and exploit new methods of agile manufacturing and rapid prototyping. This design program is based on the belief that the best designs for products are consumer initiated and have significant involvement of consumers in their design. Through the development of these new computer-integrated design tools for the quantitative assessment of function and performance of humans we are hoping to facilitate the design of customizable rehabilitation devices.

References

[1] Denavit, J., Hartenberg, H., A Kinematic Notation for Lower-Pair Mechanisms Based on Matrices, Journal of Applied Mechanics, page 215-221, June 1955.

[2] Evans, M., Magpie - It's Development and Evaluation, Internal Report, Oxford Orthopeadic Engineering Center, Nuffield Orthopeadic Center, Headington, Oxford, England OX3 7LD, 1991.

[3] Jayanthi, S., Harwin, W., Keefe, M., Kumar, V., Application of Stereolithography in the Fabrication of Rehabilitation Aids, SFF Symposium 95, University of Texas, at Austin, 1995.

[4] Mahoney, R., Phalangas, A., Consumer Evaluation of Powered Feeding Devices, Submitted to RESNA 1996.

[5] Orpwood, R., Design Methodology for Aids for the Disabled, Journal of Medical Engineering and Technology, Vol. 14, page 2-10, 1990.

[6] Wisaksana, A., Verburg, G., Naumann, S., The Development of An Interface Between the Spaceball, RESNA 95 - Proceedings, page 478-480, 1995.

JACK¿ is a registered trademark of the University of Pennsylvania.

PHANTOM¿ is a registered trademark of SensAble Devices Inc.

ProEngineer¿ is a registered trademark of Parametric Technology Corporation.

Silicon Graphics and IRIS are registered trademarks of Silicon Graphics, Inc.

Acknowledgments

This research is supported by the National Science Foundation (Grant #MIP94-20397) as well as the Rehabilitation Engineering Research Center on Rehabilitation Robotics of the National Institute on Disability and Rehabilitation Research (NIDRR) of the Department of Education (Grant #H133E30013). Additional support has been provided by the Nemours Research Programs.

Matthew T. Beitler Applied Science and Engineering Laboratories University of Delaware / A.I. duPont Institute 1600 Rockland Road, P.O. Box 269 Wilmington, DE 19899 USA

EMAIL: beitler@asel.udel.edu WWW: http://www.asel.udel.edu/~beitler Matthew T. Beitler, William S. Harwin, Richard M. Mahoney Applied Science and Engineering Laboratories (ASEL) University of Delaware/Alfred I. duPont Institute Wilmington, Delaware USA Software)