The current primary mode of operation in disaster detection and response
is through the 9-1-1 system, where humans call in emergencies and the
appropriate personnel is dispatched. This detection/response system
has proved adequate in many situations, but slow in other scenarios
(especially when the humans themselves are involved in the emergency/disaster).
Further, counting on reports from the field may be inadequate when communication
links are severed between the area affected and ``the other side.''
The goal of this proposal is to efficiently automate this process.
We propose a Self-secure and robust Critical Information Technology
Infrastructure (Secure CITI, or S-CITI for short) that takes advantage
of pre-deployed data-sensing equipment and strategically deploys new
sensors in areas of concern. For example, cameras that are used to monitor
traffic and toll booths could be redirected to report on unexpected
events (e.g., pieces falling from a tunnel ceiling or a major crash)
or re-oriented to collect information about a potential landslide in
a nearby hill.
We propose to introduce a system to coordinate the use of existing
resources and distribute the data gathered among different organizational
units. We propose a Virtual Coordination Center to facilitate
efficient and quickly coordinated actions to natural and human-caused
disasters. The system will also contain a learning module that will
feed post-emergency data into the data processing module of pre-emergency
decision making.
In order to carry out this project, we will specify scenarios and
use the information generated from these scenarios to use/redeploy sensors
for collecting specific data, establish communication between Emergency
Managers (EMs) and disseminate appropriate information. Our protocols
will respect the different organizational units and the privacy of data
through secure and authenticated data access. We will also make the
protocols sensitive to the disaster level, the load on the system and
the type of information being shared.
To manage the resources in the system (sensors, CPUs, communication
links, voice channels, cameras, etc.), we will allow EMs to specify
the importance of the data so that the system can supply the different
EMs with the most appropriate data, from a common repository. In particular,
we will dynamically synthesize databases with new consistency and authentication
procedures bound to the type, size, and importance of the data.
The vision introduced above is admittedly too ambitious a plan to
be readily implemented at the city or region levels. Therefore, we will
implement the S-CITI in three progressively more complex prototypes:
in a lab, on one floor of our Department, and in a small area in the
University of Pittsburgh campus. Lastly, we will (in the last year)
create a feasibility study for the deployment of the S-CITI within the
city of Pittsburgh, in conjunction with the Allegheny County Emergency
Services.
