Using Sensor Data to Predict the Environment in a Building.
Using Sensor Data to Predict the Environment in a
Jones, W. W.; Peacock, R. D.; Forney, G. P.; Reneke, P.
Fire Suppression and Detection Research Application
Symposium. Research and Practice: Bridging the Gap.
Proceedings. National Fire Protection Research
Foundation. February 25-27, 1998, Orlando, FL, 60-63
fire suppression; fire detection; fire research; fire
safety; fire models; predictive models; smoke movement;
As transducers become more commonplace in the built
environment, it is desirable to utilize this information
in a more complete way to assure safety. There are two
fi to doing this, incorporating our knowledge of fires
and other extreme events into the measuri and reporting
capability, and insuring that all systems are
functioning the way they were intended to. The former is
commonly referred to as smart sensing, while the latter
deals with fault detection and the needed redundancy.
These are the prime components of a system which wil
allow reliable real-time prediction of the environment
in a building. In order to accomplish these tasks, it
is important to have access to information about the
building and its environment. This requires a common
protocol to provide the data from a multitude of sensors
and sufficient computing capacity to utilize the data to
provide some indication of future events. The former
problem is exacerbated by the wide variety of sensors a
the confounding problem that sensors have traditionally
been used to signal a specific event. Tl BACnet protocol
is designed to allow a wide variety of manufactures of
sensors such as heat ar, smoke detectors to collaborate
with traditional building transducers such as velocity
probes and door-closure indicators. The protocol has the
virtue that many types of events, including analog
signals, bimodel information, reset states and many
others, exist or can be defined, allowing a great
flexibility in communication while enforcing
deterministic behavior on a potentially chaot. system.
The shortcoming in understanding what the plethora of
information means is transcende by providing sufficient
computing and memory capacity to allow reasonable
algorithms a chance to work. Taken together, we are
trying to understand what transducer actually tell us
about the environment in a building. The technical
perspective is the most fascinating. In order to predict
the environment, we must first understand the meaning of
the data that is delivered. Then we can use the
information in a system which is sufficiently faster
than real time that the predicted information is useful.