Emergency Management Through Sensor of Enterprise Systems
Emergency Management through
Sensors of Enterprise Systems
Deniz Gurkan, Kiran Vemuri, Parth Gala, Anatoliy Malishevsky,
and Anand Daga
Data sharing needs for emergency decision making
Our Proposed Integration
UH – Emergency management project
Conclusion and future work
Emergency management is time sensitive
Utilize all the information sources available to the emergency operations centers.
Data from diverse enterprise systems deployed around the campus can help speed up the
process of emergency management.
To forecast or respond to an emergency without
enough information can be as efficient as trying to
solve a puzzle without enough pieces.
Different enterprise systems and sensors deployed
around the campus can serve as information sources
for the emergency operations center to help take better
Collecting data and facilitating data exchange between
various enterprise systems and sensors by stitching them
together and intelligently generating alerts and
associations to be displayed on a common display
platform can help improve the emergency management
Collecting the data into a regular RDBMS would make the
data redundant and make it an overhead because of the time
it takes to constantly store and retrieve data.
Use of METADATA can drastically improve the time
complexity of the system as it is light, simple and fast.
IF-MAP: Facilitate Data Exchange
Defines a standard interface between any product
or system and a so called metadata access point,
which is a database for aggregating, correlating,
and distributing metadata between different
systems in real time.
Lightweight and supports three primitive
operations: publish, subscribe, and search.
Graph database implementation - simple
Client – Server architecture with sessions for every client - secure
No pre-determined schema
Run-time changes to schema – no downtime
No changes in the authoritative systems
UH – Emergency Management Portal
University of Houston’s emergency management projects aims to integrate data from systems like card reader
systems, camera systems, power systems etc., available on the campus to publish alerts to the emergency operations
center and later try to relate and perform analytics on the data to intelligently predict possible emergencies.
Use API’s if they’re open and available.
Bump in the wire
Parsing data files
Door access systems : Bump in the wire
Alerts on door access decisions ( grant/deny)
Alerts on lockdowns
Alerts on door forced open statuses
Alerts on door left open
Police dispatch : Database polling
Alerts on all the police dispatch events like burglary, weapons, rescue, etc…
Classification of alerts based on how critical the events are.
Camera systems : geographical location and access records
Buildings on the map linked to floor plans and cameras positioned in the respective floors.
Integration status cont. (Work in progress)
Facilities – alerts from fire alarms, moisture sensors/leak detection equipment etc..,
Network – alerts to indicate network failure and the devices effected.
Power – alerts to indicate power failure, areas/buildings effected and identify the cause.
Conclusion and future outlook
Incident heat map