Go to previous article
Go to next article
Return to the 1994 VR Table of Contents
By: Kenneth Nemire, Ph.D
1840 Forty-First Avenue, Suite 102
Capitola, CA 95010
408-475-8143 (voice), 408-475-8459 (fax)
Students with physical disabilities have limited access to typical science laboratories. One promising and innovative way to provide access is through virtual environment technology, which can provide a unique tool for the disabled by offering opportunities for manipulative exploration and mobility without requiring physical access, dexterity and strength. We are developing a virtual laboratory to provide greater educational opportunities to physically disabled persons. We discuss the results of our user needs and task analyses for developing a virtual laboratory for physically disabled students, and our development of classes of interactive metaphors and methods useful for future development of the virtual laboratory. Virtual laboratories will benefit disabled students by providing them greater control over their environment and more educational opportunities. The potential applications of virtual environments for education are vast, and will provide the next generation of interactive learning systems.
The typical science laboratory presents sufficient manipulation and mobility challenges that most students with physical disabilities are discouraged or prevented from participation. Our goal is to address this gap in the education of disabled students by introducing a virtual laboratory to provide easier access and opportunities for involvement and success in the sciences. In the virtual laboratory, students will be able to perform experiments to test the same concepts that would be tested in a physical classroom laboratory, without the requirements for mobility or manual strength and dexterity.
Designing the virtual laboratory, like designing any product, must first consider the users and their tasks. Our analyses helped direct selection of off-the-shelf hardware and software components for the virtual environment system, as well as provide information needed for designing new components. These analyses are presented briefly in this paper.
User Needs And Task Analyses
Analyses of user characteristics of potential users of the virtual laboratory included an assessment of the physical, psychological, and educational characteristics of potential users, including those with physical disabilities. Because it is impossible to design a system to accommodate the full range of cognitive, affective and physical disabilities, we designed the prototype virtual laboratory specifically for those with spinal cord injury. Restricting the population enabled us to address a critical segment of the disabled population, while simplifying the problems for designing the virtual laboratory. We further constrained the design problem by accommodating students who possess at least average intelligence and emotional stability, and who do not exhibit limited hearing or visual acuity. Later designs will accommodate students with a wider range of capabilities.
There are approximately 150,000 persons with spinal cord injuries living in the United States (Stover, 1986), with an estimated rate of occurrence of 30 to 32 per million per year (Parsons & Lammertee, 1991). Of all people with spinal cord injury, 36% have injuries between cervical vertebrae C-5 and C-8 (Stover, 1986). Designing virtual environments for these persons would increase the numbers who could use virtual environments, while potentially enhancing function for a wider range of users, including those considered nondisabled and those with cerebral palsy or muscular dystrophy.
The following design choices are based on estimates of the typical capabilities of persons with different levels of spinal cord injury. There is little available quantitative data to make specific design choices such as the forces and positioning of controls needed to allow use by a wider range of users.
Consequently, more human factors research is needed to provide this information. The user needs analyses were conducted by interviewing occupational therapists, assistive device specialists, and high school students with physical disabilities.
There are many physical characteristics for which one may want to specifically design a virtual environment system. These would include designing for differences of handedness, to accommodate left and right hand users, and designing for differences in strength and stature. In this paper, we will concentrate on those characteristics that may be specifically relevant for those with spinal cord injury.
The medical definition of a spinal cord injury is the site at which the injury occurred, and whether the lesion is complete or incomplete. In a complete lesion, no nerve impulses pass through the region. In an incomplete injury, impulses pass through some areas of the spinal cord and not through others. With a complete lesion, the person has no motor or sensory function below the site of the lesion; however, vision, hearing, and cognitive functions are not affected. People with a complete or incomplete lesion in the chest (thoracic or T) or lower back (lumbar or L) regions are classified as paraplegic, have impairment in the lower limbs, and probably use a wheelchair or other device to assist with mobility. However, they have full strength and control in their arms. If the lesion occurs higher on the spinal cord, in the neck (cervical or C) region, the person may lose control of the arms as well as the legs. These persons are classified as quadriplegics. Depending on the location and severity of the injury, an individual may experience impaired ability to manipulate shoulders, elbows, wrists, and fingers and thumbs.
Following are descriptions of some of the capabilities for head, arm, and hand movements in those with spinal cord injury that are relevant to use of virtual environment systems (Barker, 1989; O'Leary, 1994). These descriptions are of typical functions; function will vary from person to person. We further qualify the descriptions by assuming (1) some use of mechanical assistive devices such as tinodesis splints, and (2) no surgical operations such as tendon transfer.
C-2 to C-4: Persons with injury in these regions are able to use eye and head movements to interact in virtual environments. They do not have use of arms or hands.
C-5 lesion: Persons with injury in this region are able to control head and shoulders, and to flex elbows. The shoulders can be used to lift the arms, but the shoulders have limited endurance and often may have mild spasticity. The persons are unable to extend wrists, elbows and fingers, or to flex fingers. Arm function may be improved with the use of adaptive aids.
C-6 lesion: Persons with injury in this region are able to control head and shoulders. They also can flex their elbows and extend their wrists. There may be moderate to no spasticity in scapular, shoulder, or elbow musculature. The persons are unable to extend elbows and fingers, or to flex fingers. Arm function may be improved with the use of adaptive aids such as a tinodesis splint. This splint would allow sufficient pinch grip to hold a pen when the wrist is extended.
C-7 lesion: Persons with injury in this region are able to control their head and shoulders. They also can flex their elbows, and extend their wrists and fingers. The arms may have minimal to no spasticity, and are completely functional. However, the persons are unable to flex their fingers, and have no hand use.
C-8 lesion: Persons with injury in this region are able to control head, shoulders, arms, thumb, and first two fingers. They are unable to control the ring and small fingers. The hands and fingers may have limited strength and dexterity.
T-1 lesion: Those with spinal cord lesions in this region have full use of their hands and arms, so would not have difficulty using most of the devices in virtual environments.
Relevance of Physical Characteristics for Hardware Assessment:
The typical immersive virtual environment (VE) system consists of a head-mounted display, spatial position trackers on the head and a hand, and an instrumented glove. The VE system also may include speech synthesis and recognition systems as well as 3D auditory and tactile displays.
Off-the-shelf head-mounted displays weigh between 4.5 oz and 4.5 pounds. Few are comfortable to wear. All of them require manual adjustments to fit to the head, and some of them require manual adjustments to accommodate other characteristics of the user essential for viewing in some displays, like the distance between the eyes. Of course, an aide could make these adjustments for the student with physical disabilities, but this would not foster independence for the user. The need for control over one's environment would require a device that is comfortable to wear and can be manipulated independently by the student.
One of the difficulties with wearing a device on the head is that those with C2-C5 spinal cord lesions would be unable to raise their arms to reposition the device to avoid development of pressure spots, and subsequent pain, on their scalp or face. A person with a C-6 injury, who used a tinodesis splint, may be able to make gross adjustments, whereas those with an injury at C-7 or C-8 probably could do so unassisted.
More users would have difficulty manipulating the rotary knobs to control fit or other adjustments. With a tinodesis splint, the person with a C6 injury could use the thumb and middle finger to manipulate large rotary controls with at least three inches between them. The shoulder could be used to rotate the controls. However, the control would have to be designed to require only minimal grip forces. Those with C-7 and C-8 injury would have sufficient use of some fingers to adjust large dials requiring small amounts of force.
Even if students could operate the controls of these head-mounted displays, many displays are too heavy and cumbersome to use for periods of an hour at a time. Light-weight goggles are easier to use, and may cost less, but they generally suffer from lower resolution, smaller fields-of-view, and lack of immersion, making the virtual experience less compelling for the user.
There are several ways to accommodate the needs of those with high-level spinal cord injury. One may be boom-mounted displays (Fake Space Labs, LEEP Optics Cyberface III). These displays can be floor- or desk-mounted, alleviating the substantial muscular strain of the neck and back that occurs with long-term use (minutes to hours) of a head-mounted display. These displays also can be coupled to head or hand movements with a simple strap arrangement, still allowing most of the weight to be born by mechanical linkages. One potential drawback is the substantial cost of these displays.
Another solution may be use of large projection screens constituting the user's entire visual field (Cruz-Neira et al., 1992). While this solution has tremendous benefits because it eliminates requirements for mounting the displays on the head, the requirement for multiple projection screens and image-generators can result in a costly virtual environment system.
In the typical VE, the student can manipulate virtual objects in the virtual world by using an instrumented glove consisting of sensors that can be monitored to provide information about the angles of joints in the fingers and hand. Data from the sensors are used to animate a computer-generated image of a hand. This virtual hand follows every movement of the student's hand. Thus, the student may use the virtual hand to grasp an object in the virtual environment.
In addition, users typically navigate through the interactive, 3D virtual environment by using special hand postures that adjust the student's perspective of the virtual world so it appears as if he or she is moving in the virtual world, even though seated in a stationary chair.
Students with spinal cord lesions higher than C-7 would derive little benefit from this device. Those with injury at C-7 and C-8 would have sufficient muscle tone in their fingers and hands to don the glove, and sufficient movement to use some of their fingers. However, they may not be able to accurately reproduce various movements because of tremor and limited strength in the fingers. Those with injury at T1 would be able to use the glove with all five fingers.
Interaction in virtual environments by those with high spinal cord lesions would best be accomplished with other interaction devices.
While movement of the glove sensors overlaying the fingers result in control of the virtual hand to grasp an object in the virtual environment, a spatial tracker is used to describe the position and orientation of the whole hand so the virtual hand may move the object from one location to another, or rotate it about its axis.
Those with C2-C4 lesions would not be able to move their arms to control the actions of a virtual hand or other cursor. Some combination of head, eye, or mouth control would have to provide this function. Those with a C-4 lesion could, in addition, use voice control.
Those with C-5 lesions may be able to move their arms with their shoulders so that the hand would be extended to the sides or to head level for a maximum of ten to fifteen seconds. The shoulders would be strong enough to lift light objects for shorter periods of time. Those with lesions at C-6 level have good shoulder strength and could hold the arm all the way up or to the sides. Thus, those with C5 to C6 lesions would have sufficient control of arm movements to control a virtual representation of a hand or other virtual cursor using a spatial position tracker. These movements are used to press virtual buttons or choose items in virtual menus that provide functional equivalents to complex grasp or hand postures. In addition, other software devices are used to facilitate accurate interaction with virtual objects.
Redundant voice controls also are used to replace hand function in the virtual environment. These controls can supplement those supplied by arm movements.
Other interaction devices:
Those with injuries at C-2 to C-8 may use devices like a Spaceball, joystick, 3D mouse, chordic, or keypad device with different assistive devices and various levels of success. As with using the spatial position tracker, software enhancements in the virtual environment are used to facilitate use.
Those with spinal cord injuries at C-4 and below have sufficient control of the voice to use speech recognition systems. However, the diaphragm used for breathing is partially involved in higher level injuries, resulting in little success with using current speech recognition systems.
Psychological and Educational Factors:
Designing the virtual laboratory also included accommodation for the psychological and educational characteristics of the potential users.
The virtual laboratory was designed for high school and college students. The educational and psychological requirements for the virtual laboratory do not differ for non-disabled and physically disabled students with at least average intelligence. As expected, these students are highly motivated to learn to use a computer system, especially if it can immerse them in a virtual environment, and if it helps them perform schoolwork more effectively. However, reading and typing skills are variable, similar to those in the general population, so these variations in skill level must be accommodated in the virtual environment interface.
While the students are highly motivated and generally have a positive attitude, the virtual environment interface should be consistent, predictable, and simple to understand and operate so the students may experience control, mastery, and achievement. This will increase motivation and encourage better performance.
Part of designing a predictable system includes designing a system that is similar to what they expect for other desktop and notebook computers. Consequently, we have attempted to maintain some of the more consistent and user-friendly aspects of these interfaces.
The student also is provided many prompts in the system. These prompts are essential to help the student learn the system as well as to help learn the instructional material.
An effective method of presenting the instructional material also is essential for effective use of the virtual laboratory. We asked the students what they would want to do in a virtual environment. They all said they want to do the "extreme" physical things they otherwise may not be able to do. These "extremes" included potentially dangerous activities such as sky-diving, bungie-jumping, race-car driving, and planetary exploration. These students also wanted to play sports like soccer and volleyball, and to travel to other countries, like Mexico, to learn a new language, buy tacos from street vendors, and go to the beach. We decided it would be advantageous to design the virtual laboratory to accommodate these wishes by providing the thrills they sought with the information required from an instructional tool.
As actors in the virtual environment, students may be represented any way they like. Most wanted to choose from an array of cursors or characters. Some wanted to create their own representation. Few wanted to be represented in the virtual world as they are in the physical world.
Job and Task Analyses:
Following analyses of the user characteristics, we analyzed the tasks the students would perform. These analyses included a study of the patterns of work flow and potentially useful conceptual frameworks, or metaphors, students may use and expect in the virtual laboratory.
Usage of the virtual laboratory would depend on its availability. Currently, students may use a computer as little as 15 minutes a day because there are so few computers and so many students. Our interviews indicated that students would use the computer more often if time were made available. Data show that in the last several years, more schools are obtaining more personal computers. Consequently, it will become more likely that students can use the virtual laboratory more frequently.
The virtual laboratory is intended to supplement some aspects of the typical classroom experience and to replace others. To be successful, it may be useful to exploit the familiar experiences of the potential user and reduce the learning curve. Consequently, we use the physical laboratory as an initial metaphor for activities in the virtual environment. However, we extend the metaphor from the concrete to the abstract by employing an "exploration" or "adventure" metaphor for actions in the virtual environment. This exploration metaphor not only helps provide the thrills sought by the student with physical disabilities, but also encourages, we hope, the sense of adventure and exploration that can accompany scientific discovery.
The task analyses also included a study of the jobs or tasks the students perform in a typical laboratory. The higher level tasks include listening to lectures, taking notes, obtaining materials for the experiments, performing experiments, and discussing the procedures and results with lab partners and teachers. Various tools are used to accomplish these tasks. These include writing implements, books, calculators, and various experimental apparatuses. These tasks can be decomposed into subcomponents such as navigation, grasping, object translation and rotation, and data entry. Such decomposition allowed us to substitute other forms of interaction, in software, while maintaining functional equivalence of the activities.
This material is based upon work supported by the National Science Foundation under award number III-9361888. We thank Barbara Goldberg, Lauren Leff, Shannon McCord, Paul Mortola, Stephanie O'Leary, and the many students who generously contributed their time and knowledge. Any opinions, findings, and conclusions or recommendations expressed in this publication are those of the author and do not necessarily reflect the views of the National Science Foundation or of the consultants.
Go to previous article
Go to next article
Return to the 1994 VR Table of Contents
Return to the Table of Proceedings