Life can be
thought of as a sequence of experiences involving interactions between people
and objects. The interaction of vision, for example, is realized by the
reception and interpretation of energy packets (light) radiated by illuminated
matter. Similar interactions produce the sensory communications of speech,
hearing, and touch. These interactions can be two-way events. Humans and
animals can generate signals and movements that affect objects and other life.
However, humans alone build devices that extend in distance and in time our
ability to communicate and influence people and nature. For example,
telephones, libraries, and earth-moving equipment are a few of the machines and
systems we have constructed to extend our will to others and mold the
environment to suit our needs.
Interactions are facilitated
by interfaces. Interfaces can be either biological, such as ears and vocal
chords, or mechanical devices, such as loudspeakers and keyboards. Some -
vehicles, telescopes, television, and public address systems - extend our
natural abilities over great distances. Others - books, monuments, and
recording mechanisms - enable us to project our presence through time. Our
daily existence depends on the smooth operation of devices we take for granted.
How many of us would be able to survive without cars, televisions, telephones,
microwave ovens, computers, and information networks? These are just some of
the devices that able-bodied people interact with and rely on to get through an
average day.
|
|
Because
people with disabilities may have a diminished ability to use these interfaces,
they may experience difficulty in participating fully in daily life. In many
cases, however, the structure of our systems defines whether a person is able
or disabled. For example, if stairs were never invented, the inability to climb
stairs would not be a disability. Since society places the ability to
communicate in high regard, nonvocal or hearing-impaired individuals can
experience problems interacting with others.
Some people with disabilities
cannot produce normal "output" functions in terms of mobility and
production of communication for both person-to-person and person-to-machine
interactions; others have difficulty with the "input" sensory
functions of hearing and sight.
|
|
To cope with this problem, current devices (machines)
for people with disabilities enhance the interaction between the user (human)
and the environment (nature) and other people by providing either augmented or
alternate communication pathways. To perform mobility, communication, and daily
living tasks independently, individuals with severe disabilities must find
interface pathways to replace those that have been lost or amplify those that
are functional. High-level quadriplegics especially must overcome the
difficulty of replacing lost or diminished pathways to the outside world since
many of them can only control the muscles in their neck and above. The power to
promote user interface with a machine, translate sensor input into machine
activation, and produce a result that reflects on the environment is available
in many devices for people with disabilities.
Here are two examples of such
interfaces, drawn from my work at the Rehabilitation Research and Development
(RR&D) Center at the VA Medical Center in Palo Alto, CA.
Ultrasonic Head Control Unit
The Ultrasonic Head Control
Unit (UHCU) is an interface that allows quadriplegics to communicate their will
to the environment by enhancing their control over equipment such as
wheelchairs and specialized communication systems in a socially acceptable and
aesthetically pleasing manner.
|
The unit translates changing head positions into control
signals that operate devices to which it is attached. This design uses two
ultrasonic transducers. These transducers emit inaudible sound waves that
propagate through the air until reflected by an object. A portion of the echo
signal returns to the transmitting sensor and is detected by an electronic
circuit. The time from transmission of the ultrasound pulse to the reception of
the echo is proportional to the round-trip distance from the sensor to the
object. In the rehabilitation application, sensors are directed at the user's
head, from either the front or the rear, and on each side of the head. If the
user is in a wheelchair, the sensors are generally mounted on the back of the
wheel- chair. The sensors can also be mounted from the front. For example, if
the user is operating a computer, the sensors can be mounted on the
monitor.
The distance of each sensor to
the head and the fixed separation of the sensors describe an imaginary triangle
whose vertices are the two stationary sensors and the user's moving head. This
geometric relationship allows the offset of the apex (the head) from the
baseline and centerline of the two sensors to be calculated. The user's head
position can then be mapped onto a two-dimensional control space.
UHCU users merely tilt their
head off the vertical axis in the forward/backward or left/right directions.
Their changing head positions produce signals identical to those from a
proportional joystick. The UHCU can be thought of as a joystick substitute for
controlling an electric wheelchair, a communication aid, or a video
game.
Users of a modified electric
wheelchair equipped with a UHCU can navigate the chair by tilting their head
off the vertical axis. The changing head position is translated by the on-board
computer into speed and direction signals for the electric motors on the chair,
thus directing the motion of the chair. To travel forward, the user moves the
head forward of its normal, relaxed vertical position. Similar movements
perform the designed motion in the remaining three directions - left turn,
right turn, and backwards. This system accepts combinations and degrees of
these motions, so a smooth right turn can be accomplished by positioning the
head slightly forward and to the right. In effect, the user's head becomes a
substitute for the joystick control found on some electric wheelchairs.
Since the UHCU signals exactly
mimic those produced by a joystick, in wheelchair applications the UHCU can be
simply plugged into the motor controller. In this manner, no modifications to
the motor controller are necessary; the UHCU becomes an electronic module
providing head position control. Other devices that normally use a joystick or
switch closure as the human input mechanism can instead use an adapted
UHCU.
|
The main advantage of this
type of hardware interface is that no physical contact between the sensors and
the user's head is required. This effectively separates the user from the
device being controlled. Therefore, with this unit users should not feel
"wired-up" or confined by an apparatus around the face or body, as
frequently occurs with other interfaces. A UHCU implemented on an electric
wheelchair also has aesthetic advantages over other man/machine interfaces used
for this purpose. It seems to be more socially acceptable than alternative
designs.
In actual operation, the UHCU
wheelchair system performs satisfactorily. After about one hour of training and
practice, this system can be mastered by any individual who retains good head
position control. The head tilting required is so slight, only an inch or two,
that observers frequently cannot deduce the method of control. Since the UHCU
only responds to head tilts, the user can freely move the eyes or rotate the
head without affecting the navigation path. In this manner, the user can watch
for automobiles at intersections or converse with others while
traveling.
The UHCU is, for all practical
purposes, transparent to the operator, and the existence of any computer
hardware or software is not apparent. One "test pilot" commented that
the system was so high-tech that it appeared low-tech.
Dexter-II
Dexter-II is a
second-generation, computer-operated, electro-mechanical, fingerspelling hand.
It offers an improved solution to the communication problems that deaf/blind
people experience. This device translates incoming serial ASCII (a computer
code representing the letters and numbers) text into movements of a mechanical
hand. Dexter-II's finger movements are felt by the deaf/blind user and then
interpreted as the finger-spelling equivalents of the letters that comprise a
message. It enables a deaf/blind user to receive finger-spelled messages in
response to keyboard input during person-to-person communication, as well as
gain access to other sources of information.
This interface is designed to
allow deaf/blind users to independently receive information from a variety of
sources, including face-to-face conversation with people who do not know
finger-spelling, telephone communication, and computer access. The enhanced
communication capability will considerably improve the vocational and
recreational opportunities available to the deaf/blind community.
|
|
In operation, a message is typed on a keyboard by an
able-bodied person. Dexter-II's computer software matches each letter's ASCII
value with a memory array of stored control values. These data program
pulse-width modulation chips to operate its eight DC servo motors. Wire cables
anchored at the mechanical hand's fingertips and wound around pulleys serve as
the fingers' "tendons." As the motor shafts are energized, they turn
the pulleys, pull on the cables, and flex the fingers. These resultant
coordinated finger movements and hand positions are then felt by the deaf/blind
communicator and interpreted as letters of a message. Although the mechanical
hand cannot mimic the human hand in rotating the wrist to finger-spell a
"J," the fact that it always produces the same motions for a given
letter enhances its intelligibility.
Since it works with
electronically transmitted information, Dexter-II can be connected to a
computer, constituting an accessible "display." It can also be
operated over the telephone either from a remote computer or by a caller using
a telecommunication device for the deaf. When interfaced with a modified
decoder for the deaf, Dexter-II gives deaf/blind individuals the ability to
receive news, information, and entertainment from closed-caption television
programs.
Reactions to Dexter-II have
been enthusiastic, positive, and at times highly emotional. The increased
communication capability and ability to "talk" directly with people
other than interpreters are powerful motivations for using this interface. It
has the potential to provide deaf/blind users with untiring personal,
finger-spelling communication at rates approaching those of a human
interpreter.
The Human-Machine Integration
Section within RR&D is devoted to projects that help people with
disabilities convey their desires to their environment and facilitate
communication with others.
David L. Jaffe is a research
biomedical engineer at the Rehabilitation Research and Development Center at
the Veterans Affairs Medical Center in Palo Alto, CA
Veterans Affairs Health Care System
Rehabilitation Research and Development Center
3801 Miranda Ave, Mail Stop 153
Palo Alto, CA 94304
|