707 First Responder Support System

When

first responders enter an endangered building anything could be waiting for them. Exits may be un-findable. Parts of the structure may collapse. Hazardous materials may burn, poison, and maim victim and rescuer alike. By the time a responder reaches the building, it is already too late to do anything. The needed information: blueprints, piping and instrumentation diagrams, escape routing diagrams, structural analyses, hazmat reports, etc. are unavailable.

What we propose to do is to build a system that provides support to first responders such that the risk to them is minimized, by:
1. Helping the first responder avoid dangerous areas or situations
2. Helping the first responder to perform his or her duties as quickly and effectively as possible
3. Providing the first responder with the safest exit alternative at all times.

Cost estimates derived from the prototype we have built suggest that we could construct a preliminary version of the systems with base funtionality in the range of hundreds of dollars per building floor and hundreds of dollars per first responder. These costs will naturally drop in the future as wider deployment creates economies of scale and devices that are more powerful, less expensive, etc.

The System

707 is a joint effort between the Usability Study Laboratory and The Visual Information Systems Center of the School of Information Sciences at the University of Pittsburgh. At the heart of the system is a computer generated three-dimensional architectural model of the building or facility where the emergency exits and other important features are noted. This model is constructed from scans of the individual floor plans for the building.

The digital models of the floors are connected together at the appropriate points to form a three-dimensional wire-frame digital model of the building. Next, digital photographs taken on each of the floors are "painted" on the wire frames of the floor. The result is a fairly accurate digital visualization of the entire building. The actual prototype model was generated from sample images such as the one shown below.

The model is augmented with a database of information about hazardous materials, structural weaknesses, escape routes, etc. The image view is automatically annotated with colors and or icons denoting important information about the building (e.g. are there hazardous materials inside?; does this door lead to an exit?; etc.)

The building model would be augmented with a "vicinity" model that would include the street access to the building and any other important information. These models could be easily generated from aerial photographs. Most cities have commissioned aerial surveys and the US Geological Survey (USGS) produces Digital Ortho Quarter Quads (DOQQs) which are readily available at no cost.

Dispatch & Command

In planning and directing emergency responders it is crucial to develop an accurate picture of damage to the structure, responder locations and activities, and probable locations and situations of victims. A shared 3-D model would be an ideal tool for dispatchers to mark up locations of fires, collapses, and probable victims.

Training

While on-site training exercises provide responders with direct exposure to the buildings they may be called upon to save, such exercises are expensive, disrupt the normal activities of the occupants, and are limited to the parts of the building involved in the exercise scenario. Training based on computer generated 3-D models by contrast is inexpensive, non-disruptive, can cover the entire facility, and can be conducted repeatedly at the student’s own pace. In other ways a virtual environment is superior to a physical one.