Graduate students from the Occupational Therapy Department at Saginaw Valley State University do the research for this section of the proposal. They divide into research teams and work with Doug Baldwin and Janet Nagayda to produce the valuable contributions below. These students are the pioneers who did the initial studies that were (eventually) presented for consideration to the leadership at SVSU.

Special thanks to:

Winter and Spring 2007:

1. Julie Ritsema, OTS SVSU
2. Julie Hadden, OTS SVSU
3. Danielle Scott, OTS SVSU

Spring and Summer 2007:

1. OTS SVSU
2. OTS SVSU
3. OTS SVSU

Winter and Spring 2007:

1. Jennifer Krieger, OTS SVSU
2. Nicole Nikolai, OTS SVSU
3. Kristi Prieur, OTS SVSU

1. OTS SVSU
2. OTS SVSU
3. OTS SVSU

As I met with the students listed above, there was always a unanimous consensus that it would be easy to make the case that Occupational Therapists should play the primary leadership role in the AT program. For this reason, I put the separate link below where the graduate students can put their rationale/ justification for OT leadership:

This is an introductory outline of potential research and development goals. The following areas of AT research have been singled out as innovative and undeveloped. They argue for the creation of CAT.

Smart Spaces: Bringing Special Education Buildings into the New Century: This is a huge concept with many variables. One of the major products in a pipeline would be what we call "Windows" ("Magic Windows", "worm holes"; we use several names to describe the concept). Magic windows allow the user to surf "windows" that are in another part of the world. They are smart windows (they look like windows in a wall) but the view can be changed with a remote. The window is linked to live feed windows that give real time views of (for example): the Chicago skyline from the 100th floor of the Sears building; a virtual window on the moon looking back at the earth; a window into 18th century Victorian London; a window into your daughters preschool room; a window out the porthole of a space ship on the way to Mars; and so on. Another way to think about Magic Windows is to envision Snow White in front of her magic mirror "Mirror, mirror, on the wall, who's the fairest of them all?"; we will give voice commands to the Magic Mirror. Only our Imagination can set the limit to what we can do with these virtual images.

Spaces are going to get very, very smart. A meshwork of smart stuff will communicate and cooperate without human knowledge or control. This complex network will have the following possible combinations:

01. WAN: Wide Area Networks like the internet

02. LAN: Local Area Networks, like agency or government personnel on a common communications channel

03. PAN: Personal Area Networks, like wearable computers: Clothing will become smart and will monitor and assist with personal needs. This has vast implications for handicapped kids.

04. IAN: Internal Area Networks; computer chips embedded inside living things: This is where medicine merges with Special Education and Rehabilitation.

05. SAN: Spatial Area Networks; smart spaces that communicate with each other and with smart IAN's and PAN's; ie intersections, roadways, the inside of cars. A Sub-network here would be the AAN (Acoustic Area Net; the linking of audified nodes)

06. OAN: Object Area Networks: things that communicate with other things: car parts that communicate; wheelchairs parts that self diagnose and communicate with the repair shop; Signs that talk and are smart enough to recognize people and other relevant objects.

07. VAN: Virtual Area Networks: GPS locations that hang in space anywhere and that communicate; ie information in places.

08. MAN: Molecular Area Networks: the meeting and networking of biology and machines

09. NAN: Nano Area Networks, chemistry meets machine

10. QAN: Quantum Area Networks: machines and light at the level of God

A blind student, wearing smart clothing (a PAN; personal area network) approaches a busy intersection. The intersection is part of the federal government's intelligent highways network; it is a smart intersection (SAN, a spatial area network). Many of the cars that pass through the intersection are smart cars, they have LAN (local area networks; car specific computer systems) and OAN systems- they can communicate with the intersection and with the blind persons smart clothing. The light changes as the blind individual arrives. A message alert is sent to the smart vehicles ("blind individual crossing intersection"). The smart cane that the pedestrian carries is guided by the intersection so that the blind person makes a direct crossing. When safely across, the intersection and smart cars note this and proceed on their way.

Hallways and classrooms should have embedded micro-processors that serve multiple functions: teach (allow) the children to remotely control environments; track and guide semi-autonomous vehicles; alter lighting to suit the needs of individual kids under variable circumstances; provide objects and landmarks in a smart space that talk and converse with children; and so on. Children should wear smart clothing and drive smart vehicles that tell the environment who the kids are and what their specific circumstances require. Environments also need to be modified to adjust to specific disabilities. We can now make the world more visible for the blind, but only if we embed smart technologies into the fabric of the environment.

Digital Vision for Children in Special Education: Digital Vision will eventually change what it means to see; human perception will be altered.

An interview about Dr. Steve Mann and his invention explains how digital vision has become a reality.

Digital Vision will change what it means to be an eye doctor; new professions will be created.

This technology will be useful for various consumers, including people who are blind, visually impaired, autistic, deaf, deaf blind, physically impaired, learning disabled, people with memory deficits, and cognitively impaired individuals. It will also have a huge carry over to people who are sighted who want to use bionic perception (multi-sensory).

The baseline technology works like this:

A digital camera (nearly or completely invisible) captures what the eyes are seeing (in real time) and presents this image to the retinas (the photo cells at the back of the eyes). In other words, a person wearing digital vision spectacles sees the world normally, as anyone else would see the world; the difference is that the real image is replaced (simultaneously) by a laser image placed directly on the inside surface of the eyes.

The significance of this is that the camera image can be digitally altered in whatever way software and hardware allow. Let's consider some ways we might digitally alter the images that the eyes capture in real time:

1. We can present to one or both eyes, in a complete field or in quadrants of the visual field, images from one or more cameras. Therefore, if we place cameras in a headband, we can see 360 degrees or in any part of the global field that we wish- rear view vision, for example. We can also "look" out remote eyes, like out the eyes of robots, or nanobots (moving through the blood stream), or through cameras placed in the environment, or through live feed video of famous locations (the acropolis, for example), or into the classroom of our preschoolers, or- through a system Steve Mann calls "seeing eye people"- out the eyes of others (your teenage friends, your college student on vacation, etc.). Mobility specialists can look out the eyes of their visually impaired clients.

2. Digital software will theoretically (eventually) look at a page of print (or a sign at a distance) and pick the figure out of the ground, enhancing the figure using contrast algorithms, and/or enlarge the image digitally. The ground can also be suppressed or eliminated. Ocular character recognition will also be used in a head-mount for blind users. Object and face recognition will be used for identification of people, animals, and objects. Gesture and body language (facial emotion, for example) software will also be used to help identify unspoken communication.

3. Digital vision will allow psychologists to experiment in unusual and important ways. Eye movements (where people look) could be constantly monitored (live, by as many experts as want to look, and/or be digitally recorded). We could determine what happens when we give a monocular eye rapidly alternating binocular images- self adjusted. We could feed images in real time (over whatever time frame we selected) to the right or left hemispheres (while inhibiting the other); we could explore the effect of long term hemispheric amblyopia. We could selectively block peripheral input. We could change processing speed (image presentation speed). We could experiment with increasing or decreasing contrast, color, processing presentation, flashing figure against the ground, etc.

4. Digital vision can use other energy levels of the electro-magnetic spectrum. This would allow seeing with ultraviolet vision, infrared vision, penetrating radar vision, as well as standard polarized vision, and various sun-blocking prescriptions. It will also allow for smooth acuity adjustment as the eyes look at various distances (correcting for presbyopia, night vision, and developmental changes in refraction). These functions would be under user control- self adjusted or tuned. Using various electromagnetic spectrums with computer algorithms will eventually result in seeing through smoke, walls, and water depths.

Autonomous Vehicles for Blind and Visually Impaired Children: For a layman, the image of a blind child driving a car (no matter how small the vehicle) is unbelievable. Even well educated futurists might roll their eyes at the suggestion that the first task of a center for assistive technology would be to produce a vehicle that could drive itself.

Yet consider the reality:

In 2005, five autonomous vehicles raced across the Mohave Desert, completing the 200 mile journey without mishap. This race held a 2 million dollar top prize and it was sponsored by DARPA, the nations think thank for the military.

In 2005, the National Federation of the Blind, the largest consumer group of blind individuals in the world, announced a plan to sponsor the "Blind Drivers Challenge". Inventors are encouraged to create a semi-autonomous vehicle that can be controlled by the blind. The contest is sponsored by several major organizations, including: General Motors; Microsoft; NASA; DARPA; and Carnegie Melon University (the leading robotics program in the nation)

The United States Department of Transportation has a 200 billion dollar program to design and develop an intelligent highway network. Our streets (the environment through which smart vehicles will travel) are about to become "smart spaces".

An entire new discipline has quietly evolved over the past ten years: The study and creation of "Intelligent Ground Vehicles".

I have a picture from 1980, taken in the gym at a basketball game. Both sides of the gym are crowded with spectators, children in their silver manual wheelchairs, with their attendants ("wheelchair pushers") lined up behind them. It looks like a snapshot from a third world country.

If you visit the hallways of Millet now you see beautiful three year old physically impaired kids driving their fire engine red, high tech power chairs. We have to teach drivers training these days; only the most severely cognitively impaired students are ever pushed in those old silver chairs.

The evolution of wheelchair technology will continue to advance because of Moore's Law: "Every 18 months computer technology doubles processing power." This means that about every two years, like clockwork, wheelchairs are manufactured that are twice as smart as the last generation. The new chairs are also lighter, stronger, and slightly cheaper than the last incarnation.

Wheelchairs are evolving into ectoskeletons: The trend is smaller; cheaper, smarter, lighter. The components of wheelchairs will shrink and become wearable. Machines will support handicapped kids, help correct posture anomalies, and will provide for mobility. In other words, an exterior skeleton will evolve. Chairs should also have strobe and sound beacons for outdoor travel, especially when young visually impaired kids are crossing streets. Many adaptations should be available that currently are not.

Smart canes; light sabers: The trend is to put ever more functionality into ever smaller, smarter, and cheaper packages. Cell phones, i-Pods, remotes, GPS systems, and PDAs are merging. The same trend can be translated into technologies for handicapped kids. An example is the porting of technologies to the long cane used by blind kids. There is no reason we cannot add GPS, sonar, laser, tactile sensing, and much more to these canes.

Voice activated computers: Many categories of children in special education could benefit from smart computers that understand and can converse. The problem we continually face, as teachers and parents, is that these technologies are mass produced for generic populations. Consequently, they never quite fit for specific children; and they are very expensive. What we need is custom designed and continually modified smart computers.

Special Education game machines: We can modify arcade games to fit the needs of children in special education. This is a huge idea with unlimited applications. Driving games that already exist could be slowed down, tailored to individual kids, and made into teacher-controlled academic modules (answer a question correctly and get two more minutes of driving time). These machines could extend range of motion, do visual motor training, and provide virtual reality experiences (street crossing practice, for example).

Brain Control of Motor Movement for Children in Special Education: New technologies are enabling people with no hand or leg control to manipulate the environment using thoughts. "The Brain Port" developed at the University of Wisconsin at Madison is an example. We experimented with one of these technologies at Millet, but we lacked the resources and support to effectively make a difference.

Vision Therapy: Knowledge from the profession of occupational therapy must be combined with the knowledge of developmental and low vision optometrists. Add to this the knowledge base of orientation and mobility specialists (blind rehabilitation professionals) and you get a new kind of professional discipline. The marriage of these three professional disciplines (plus others?) creates an expert system that can enhance and rehabilitate (potentially?) individuals with many kinds of vision anomalies (autism to head injury trauma).

Medical meets Rehabilitation meets Special Education at the molecular level: This is another huge conceptualization. We are embedding computer chips in the body and then we are linking these chips wirelessly to each other (creating an internal area network) and to computers outside the body. We are developing stem cell solutions for specific anomalies. We are altering the human body through genetic engineering and smart drug applications. We are peering ever deeper into the human brain with our imaging technologies. We have a diagnostic arsenal that is ever more powerful as the years roll on. We could create a detailed profile of our handicapped children that goes way beyond what we are doing in this day and age, but it requires unprecedented cooperation between the medical and educational communities. These potential systems are not applied to handicapped kids because there is not enough economic incentive for big corporations to pay attention. For those of us charged with bringing the best solutions to the most impaired and disadvantaged children, there is an obligation.

Research into navigation with CTFM ultrasonic sensors A blind boy hits a softball pitched toward him with a baseball bat. He then hops onto his bicycle and rides along a path lined with cherry trees. Can we build a mobile robot to drive around the paths on a campus without running over the garden by sensing plants and path surfaces? Can we design an aerial robot to fly into a partially collapsed building to look for casualties in a disaster by sensing openings in the walls of the building? The blind boy is using an ultrasonic mobility aid to sense his environment. He has learned to navigate using echolocation. The aid continuously transmits a frequency modulated ultrasonic signal (CTFM). The echoes are demodulated with the transmitted signal to produce audio tones that are played through head phones placed near his ears. The frequency of the tones is proportional to range and the amplitude to object size. We are seeking to achieve human like navigation ability for autonomous vehicles (both ground and flying) using CTFM ultrasonic sensing. Humans use imprecise geometric information to navigate. Our hypothesis is that humans do not need precise geometric information because of their ability to accurately perceive and track landmarks.

Our goal is to use the rich information in the CTFM ultrasonic echo to autonomously navigate a vehicle. This involves perceiving objects suitable for landmarks, creating an information rich map of the environment, planning paths from that map, and navigating along those paths using landmarks to localize. Echolocation is the perception of objects and their location from the echoes of chirps of ultrasonic energy off those objects. Bats use it to navigate in the dark and in restricted spaces, such as in forests and inside buildings. It is a sense of perception that human´s don´t normally poses. If God had not make echolocating bats we would not believe it possible to recognise objects and navigate using ultrasonic sound waves. We are surprised by the ability of blind people to learn to use mobility aids based on ultrasonic sensing systems. Ultrasonic mobility aids for blind people were commercialised by Em. Prof. Leslie Kay. With them, we have developed ultrasonic sensing systems to recognise leafy plants [1] and to discriminate between ground surfaces based on their roughness [2]. This research demonstrated that the echo contains information about the geometry of the objects. It also demonstrated that the quality of plant recognition is improved by using features extracted from the echo rather than signature matching of the echoes. We were able to divide the feature set into range dependent features for range measurement and range independent features for plant classification. Also, orientation dependent features are useful for bearing estimation and orientation independent features for classification. Wheeled mobile robots travel across a variety of surfaces: concrete, linoleum, grass, etc. Bat researchers found that bats can distinguish between perches based on their surface roughness. Similarly, we have achieved excellent classification of surfaces with features extracted from CTFM echoes [2]. The sensor is held at an angle to the surface the vehicle is traversing. A moving sensor produces better results than a stationary one. A characteristic of ultrasonic sensing is that motion improves perception. In parallel with the sensing research, we have been experimenting with using CTFM ultrasonic sensing to detect landmarks for autonomous vehicle navigation [3]. Landmarks can be grouped into classes based on their complexity and their continuity. A simple discontinuous landmark is a pole in the middle of a field. A system using the "Visual Flight Rules" approach from light aircraft navigation plans a path as a set of legs between landmarks [4]. Each leg consists of a distance and bearing to travel to the next pole.

The robot was taught a map by manually driving it from one landmark to the next. Then it was placed near the start location with an initial error in both range and bearing from the map values. It used a compass and odometry to travel to the next landmark. Upon reaching the location where it expected to find the landmark it scanned for it. The range and bearing of the landmark was then used to replan the next leg. After traversing three legs the initial errors were removed. The edge of a path is a simple continuous landmark. We have navigated a mobile robot in the Wollongong botanical gardens by tracking the edges of paths [5]. The grass growing beside the concrete path formed a boundary that was easy to detect. In current research, we are developing a mobile robot to follow paths using measurement of surface roughness. Blind users of the mobility aid use a horizontal scanning motion of the sensor to sweep the beam from side to side in the direction that they are walking ro detect both the path and its edges. We are studying how humans navigate using CTFM ultrasonic sensing: the perception of objects, the scanning of the sensor to improve object detection, the effect of moving on the sensor information, and how they combine this information with their mental maps and plans to navigate to a desired destination. Our aim is to use this understanding to navigate both mobile robots and autonomous aerial vehicles.

[1] McKerrow, P.J. and Harper, N.L. 2001. ´Plant acoustic density profile model of CTFM ultrasonic sensing´, IEEE Sensors Journal. Vol 1, No 4, 2001, pp 245-255. [2] McKerrow, P.J. and Kristiansen, B.E. 2006. ´Classifying surface roughness with CTFM ultrasonic sensing´, IEEE Sensors Journal, Vol. 6. No. 5., October, pp 1267-1279. [3] Harper, N.L. and McKerrow, P.J., 2001, ´Recognising plants with ultrasonic sensing for mobile robot navigation´, Journal of Robotics and Autonomous Systems, Elsevier,Vol 34, Issue2-4, pp 71-82. [4] McKerrow, P.J., and Ratner D. 2001. "Navigating an outdoor robot with simple discontinous landmarks´, Proceedings EUROBOT´2001. Lund, Sweeden, September, pp 33-39 [5] Ratner, D. and McKerrow, P.J. 2003. Navigating an outdoor robot along continuous landmarks with ultrasonic sensing, Robotics and Autonomous Systems, Elsevier, Vol 45/2 pp 73-82.