Problem Scenario: Suppose you are blind person who works in a campus setting, i.e. a cluster of buildings all linked by various pathways containing many obstacles. If you wanted to travel from one location on campus to another a list of things to worry about (consider) such as the following may begin to grow in your head:

• How fast do I need to get there

• What is the weather like

• Should I take a sheltered route

• Which route has the fewest obstacles

• Which route has the least amount of traffic during this time of day

When you enter a room on campus, you wonder about its layout,

where the chairs are, where the black board is, where the desk is

Also you might like to know where the nearest restroom is, or nearest water fountain is

This looks like a case for a personal computing assistant, with differential GPS, and a database

of stored information about pathways, context sensitive information with respect to time, weather,

location. Much of this information can spill over into useful information for sighted people.

Problem definition: The problem is to design a system that using differential geographical positioning system (DGPS) information can precisely navigate a blind pedestrian from one location on campus to another location on campus. While also providing rich content about each location along a route, and choices in route from one location to another where the choice could be made based on traffic conditions for a particular time of day, or current weather conditions.

Investigation into a solution: First of all I need to come up with the model for route traversal, the thought here is to model a route from one location to another, as a series of locations or position cells. This way a traveler can be in a cell full of information specific to that cell and maybe interesting to the traveler in the cell. Such as if you are in a hallway and there is a railing along the wall, railings are always nice guides that can help a blind person traverse a hall with more confidence. The system should be able to keep the traveler on course through a cell, onto the next cell on the route. Keeping on route will be done with verbal prompts such as "turn 45 degrees to the right and continue forward", or "a little to the left".

The model should facilitate the building of routes by traversing them with the PCA and specifying the endpoints of the route, it should also be easy to add context sensitive information to each cell, such as varying degrees of traffic depending on day or time of day.

I will look at the possibility of incorporating some of the information that the University keeps on the state of classrooms into the systems information base.

In order to provide location accuracy to within a few centimeters, I think this accuracy is required to navigate a blind traveler through the maze of obstacles on campus, I will need to use DGPS. This is done by using a GPS system together with a master GPS system that is stationed at a carefully surveyed location. The master can provide the calibration information necessary for the travelling GPS to overcome the error injected into the GPS satellite signals.

The PCA will need to be able to provide the ability to voicify navigation help, and actually provide a verbal interface to the blind user.

Since much of the information being modeled, such as how to get to a specific location, some of the content of each cell, "how do I get to the nearest restroom" appears that it could be useful to more of the Universities population than just the blind, I will look into covering the campus with information dissemination servers. This way when a person is traversing campus they can have the necessary information for there location automatically broadcast to them. This can reduce some of the burden of the PCA carrying around so much information about the campus that it does not need all at once. Some of these information servers can also serve as master GPS servers. Of course a sighted person will probably not need to use the GPS system, so we can relax its use when not needed. The DGPS system could be turned on by a blind users profile.

If I model the system as a distributed set of information dissemination servers I could tap the system into at least the Universities network of Web content. Thus one could access library information, ISIS information, and so on from anywhere on campus.