Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
2189414
SYSTEM AND METHOD FOR ESTIMATING CHARACTERISTICS OF BROADCAST
RADIO AUDIENCES
FIELD OF THE INVENTION
The present invention relates generally to survey systems
that passively detect radio stations to which radios in
passing vehicles are tuned.
BACKGROUND OF THE INVENTION
Most automobiles and other vehicles include radio tuners
designed to receive commercially broadcast radio programming.
A few audience survey systems utilize electronic RF equipment
to passively detect and analyze car radio usage. Radio
audience survey systems typically employ an individual testing
site that gathers survey data over a predetermined time
period. The survey data may identify the most popular radio
stations, the estimated number of radio listeners, or audience
listening preferences at the particular testing site. This
testing site may be moved from time to time to gather survey
data for other sites during other time periods. The raw
survey data is typically compiled and transformed into
readable radio station ratings reports, tables, or charts.
Unfortunately, conventional radio audience survey systems
may be inadequate for certain applications. For example,
current systems that are limited to sequentially collecting
survey data at different locations within a city may be unable
to adequately estimate radio station preferences for a large
geographical region. Generally, conventional test site
equipment and operation is rather expensive. Calibration,
data collection, and data retrieval procedures typically must
be performed at each individual testing site with a certain
amount of human interaction. These conventional procedures
make survey data obtained at one site and time incompatible
with other survey data taken at another site and time.
1
2189414
The statistical accuracy of RF survey systems is
partially dependent upon the number of RF signals detected
(i.e., a greater amount of data leads to greater statistical
accuracy). The amount of data collected by a passive system
may be dependent upon the speed of the vehicles as they pass a
testing site, the scanning speed of the electronic equipment,
or the number of test sites located within the desired
geographical area during a survey period. Due to a limited
ability to collect survey data in a cost efficient manner,
known survey systems may only be capable of estimating radio
audience preferences on an hourly or daily basis. Thus,
conventional RF survey systems may be inadequate if survey
data is desired over shorter time periods, i.e., a few minutes
or less. An improved time resolution is needed when
information describing listener responses to short programs,
commercials, and/or songs is desired.
Conventional radio audience survey systems provide
statistical reports for radio programming broadcast in the
distant past. These reports are often made available weeks,
or even months, after the radio programming has aired. The
reports typically provide little or no information describing
what was broadcast by different radio stations during the time
period when the survey data was collected. Thus, radio
station managers, advertisers, and music industry
representatives are provided with no direct listening audience
feedback in response to specific radio programming.
SUMMARY OF THE INVENTION
Accordingly, it is an advantage of the present invention
that an improved system and method for estimating
characteristics of broadcast radio audiences is provided.
Another advantage of the present invention is that a
large quantity of survey data based upon vehicle radio usage
may be collected in large or small geographical areas.
2
2189414
Another advantage is that a system according to the
present invention is capable of collecting a large amount of
survey data without utilizing an impractical amount of human
labor.
A further advantage of the present invention is that it
provides a passive radio survey system capable of detecting
radio audience preferences over relatively short time
intervals.
Another advantage is that the present invention provides
radio survey data associated with radio programming broadcast
during the time period that the survey data is collected.
An additional advantage of the present invention is that
radio survey results are made available shortly after the
surveyed time period.
The above and other advantages of the present invention
are carried out in one form by a method for estimating
listening audience preferences in response to radio broadcast
programming. The method involves monitoring a radio program
broadcast by a radio station during a time period, collecting
survey data describing radio stations to which tuners are
tuned during the time period, and associating the radio
program with the survey data in a substantially synchronous
format.
BRIEF DESCRIPTION OF THE FIGURES
A more complete understanding of the present invention
may be derived by referring to the detailed description and
claims when considered in connection with the Figures, wherein
like reference numbers refer to similar items throughout the
Figures, and:
FIG. 1 is a block diagram representation of an audience
survey system according to the present invention;
FIG. 2 is flow diagram of a broadcast/survey associating
process performed by the audience survey system;
3
2189414
FIG. 3 is a flow diagram of a survey data collection
procedure performed by the audience survey system; and
FIG. 4 is a flow diagram of a calibration process
performed by the audience survey system.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
FIG. 1 illustrates an exemplary representation of an
audience survey system 10 in accordance with the preferred
embodiment of the present invention. In general, survey
system 10 includes a central office 12, a multiplicity of
remote survey sites 14, and a plurality of remote access sites
16. Survey sites 14 and access sites 16 are in data
communication with central office 12: Survey sites 14 are
distributed throughout a geographical area, such as a city, a
county, a state, or the like. Although not a requirement of
the present invention, survey system 10 may include an
additional number of central offices (not shown) combined
through a data communication network to support larger
geographical areas. Although the preferred embodiment is
shown and described as a system for estimating listening
audience preferences in response to radio broadcast
programming, the present invention may also be adapted for use
as a system for estimating other characteristics
representative of a population of vehicles. Thus, the present
invention may be utilized with traffic control systems, car
counting systems, and any number of systems that utilize
vehicular survey data derived from the detection of radio
tuner local oscillator emissions.
Survey sites 14 are each configured to passively collect
survey data describing radio stations to which tuners are
tuned. According to the preferred embodiment, survey sites 14
are located proximate a road or highway (not shown) to enable
the electronic detection of vehicle radio usage. To enable
the collection of a statistically relevant amount of data, a
number of survey sites 14 are preferably located throughout a
4
2189414
geographical area. For example, survey sites 14 may be
located throughout many major metropolitan cities to collect
survey data for nationwide or regional analyses.
Alternatively, survey sites 14 may be~densely located within a
city to provide inter-city traffic or radio audience survey
data. The parallel accumulation of data through many survey
sites 14 enables survey system 10 to collect and process
survey data for relatively short time intervals (several
minutes or less).
Each survey site 14 includes a receiver 18 and a control
unit 20. For clarity, standard operating components known to
those skilled in this art are not shown or described herein.
Radio tuners generate local oscillator (LO) signals for the
demodulation of received radio signals,- and a portion of each
LO signal is emitted (rather weakly) from the tuners through
the vehicle antennas. Thus, receiver 18 employs an RF antenna
22 to detect the~LO signal emissions (which are associated
with the transmitting radio station frequencies) from vehicles
passing survey site 14.
Receiver 18 is calibrated with a site-specific LO
detection threshold such that LO signals are selectively
detected. In other words, many LO signals which otherwise
have sufficient signal strengths to be detected by receiver 18
are ignored. The calibration of the threshold is controlled
remotely by central office 12, and the threshold may be
remotely adjusted from central office 12 (described below).
The calibration procedure, receiver control functions, and
other processes related to the present invention may be
controlled by control unit 20. According to the preferred
embodiment, survey data describing radio stations to which
tuners are tuned may be accumulated or formatted at control
unit 20 before being downloaded to central office 12.
Central office 12 functions as a processing center for
survey system 10 and is preferably located at a remote
location relative to survey sites 14. Central office 12
preferably controls the operation of survey sites 14, receives
5
2189414
survey data from survey sites 14, and transmits processed
survey data to access sites 16. Central office 12 includes a
radio program monitor 24, a computer 26, and a system
controller 28. Computer 26 includes .at least a processor 30
and a memory 32. Again, for clarity, specific computer
components well known to those skilled in the art are not
shown in connection with computer 26. Processor 30 is
configured to perform various operating processes described
below and is connected to memory 32, which stores downloaded
data from survey site 14 and programming instructions that
define the various operating processes described herein.
Radio program monitor 24 is utilized to monitor a
plurality of radio programs broadcast by a corresponding
plurality of radio stations during the..same time that the
survey data is collected. FIG. 1 depicts radio programming 23
broadcast by transmitting antennas 25 corresponding to
different radio stations. Radio program monitor 24 desirably
time stamps and digitizes the radio programming for storage in
memory 32 or for further processing by processor 30.
According to the preferred embodiment, some or all of the
radio stations in a city are concurrently monitored. Thus,
survey system 10 is capable of producing a record of the radio
programming broadcast by those stations during particular time
periods.
System controller 28 may be a separate user interface or
an integral part of computer 26. System controller 28 is
preferably utilized to enable system operators to remotely
control various processes described herein. For example,
system controller 28 may allow an operator to remotely adjust
the LO detection threshold utilized by survey sites 14 or to
remotely instruct survey sites 14 to enter a calibration mode
(described below).
Access sites 16 may be configured to facilitate online
access to central office 12. According to the preferred
embodiment, access sites 16 communicate with central office 12
through either cellular or wired telephone modems. The
6
2189414
present invention may alternatively utilize any suitable
remote data communication device. Thus, access sites 16 may
be mobile, temporary, or permanent stations. Access sites 16
include a data presentation device 34 .having an audio output
S 36 and a video output 38. Data presentation device 34 is
preferably connected to a user interface 40. Data
presentation device 34 may, for example, be a conventional
personal computer having as much processing and memory
capacity as needed. As such, audio output 36 may be
implemented using a conventional personal computer sound card
and speakers, and video output 38 may be implemented using a
conventional video monitor. Data presentation device 34 is
configured to provide an audio representation of radio
programming while displaying a visual representation of the
collected survey data. User interface 40 is preferably
configured to enable an operator to interactively manipulate
the audio or video data.
Referring now to FIG. 2, a broadcast/survey association
process 42 performed by survey system 10 is depicted as a flow
diagram. Process 42 preferably begins at the beginning of a
time period during which survey data will be collected. For
purposes of this description, "time periods" are used to
conveniently describe the operation of process 42, and process
42 preferably does not operate with discrete blocks of time.
In addition, a "radio program" may be any portion of a
particular radio station broadcast and need not be a specific
song, commercial, or other temporally defined segment.
Furthermore, nothing prevents the preferred embodiment from
continuously monitoring radio programming and continuously
collecting survey data.
Process 42 begins by performing a task 44 and a procedure
46 substantially simultaneously. Task 44 monitors radio
programming broadcast by a plurality of radio stations. The
radio programming of each radio station may include any number
of individual radio programs (such as songs, commercials, news
briefs, and the like). Task 44 is performed,by radio program
7
2a~9~1~
monitor 24 (see FIG. 1). Using known techniques, task 44 may
add time and date stamps while monitoring the radio
programming and digitize the monitored radio programming for
storage in memory 32. Survey system 10 eventually associates
S an audio representation of the monitored radio programming
with a visual representation of the collected survey data.
Because a high fidelity reproduction of the radio programming
is not required by the present invention, task 44 need not
fully record the radio programming.- For example, task 44 may
only sample the radio programming or digitize a limited
bandwidth representation of the radio programming. In an
alternate embodiment, task 44 may monitor the radio
programming and associate an identification with the radio
programming. While task 44 is being performed, procedure 46
collects survey data describing radio stations to which tuners
are tuned during the time period. Procedure 46, which is
performed at survey site 14, is described below.
Task 44 and procedure 46 eventually lead to a task 48.
Task 48 downloads the survey data collected during procedure
46 from survey site 14 to central office 12. In addition,
radio program monitor 24 may download corresponding radio
programming data to computer 26 (see FIG. 1). Preferably,
task 48 utilizes a telephone modem or other data communication
techniques to download the survey data to central office 12.
If desired, task 48 may be performed at substantially the same
time as procedure 46 to achieve realtime data acquisition and
processing. Alternatively, a short time delay may be
implemented by survey system 10 such that survey data
collected in the recent past is downloaded by task 48 while
task 44 and procedure 46 continue to operate. Nothing
prevents task 44 and procedure 46 from running continuously,
while task 48 downloads data only when necessary. Survey
system 10 may be configured such that survey sites 14
automatically place periodic downloading calls to central
office 12 for downloading of survey data, or vice versa.
8
2189414
After task 48 downloads the survey data to central office
12, a task 50 processes and compiles the survey data according
to specific system requirements. For example, task 50 may
relate detected LO frequencies collected in the survey data to
the corresponding transmitting frequencies and/or radio
station call letters. Task 50 may also format the survey data
by geographical location, time of day, or other categories.
In addition, task 50 may process data related to other
characteristics representative of the vehicles traveling
within survey system 10, such as the average speed of the
vehicles, the traffic density in specific areas, or the number
of vehicles passing survey sites 14. Those skilled in this
art will recognize that task 50 need not be performed at
central office 12, e.g., task 50 may be performed at survey
sites 14 before the survey data is downloaded or at other
times during broadcast/survey association process 42.
After task 50, a query task 52 determines whether an
access request is detected. The access request instructs
survey system 10 to generate a report (not shown) containing
formatted and compiled survey data. The access request may be
input by any number of users located at access sites 16 or by
a system operator at central office 12. Alternatively, the
access request may be an internal parameter of survey system
10 that periodically prompts central office 12 to generate a
survey report. If query task 52 determines that no access
request is detected, then broadcast/survey association process
42 ends. Process 42 may exit into a standby mode or continue
to monitor radio broadcasts and collect survey data. If query
task 52 determines that an access request is detected, then a
task 54 is initiated.
Task 54 gets a set of user parameters that describe how
the survey data is to be formatted. According to the
preferred embodiment, access site 16 is online with central
office 12 during task 54. The user inputs his or her format
preferences at access site 16, and central office 12 responds
accordingly. For example, a user may control the time window
9
2189414
subject to analysis, the radio programming segment to be
heard, the radio station to be analyzed, and/or the selection
of different demographic databases. Conventional data base
manipulation techniques may be implemented by survey system 10
during task 54.
After task 54, a task 56 associates the specified radio
programming with the survey data in a substantially
synchronous format. For purposes of this description,
"synchronous" refers to the time relationship between the
radio program and the survey data. For example, if a radio
station is to be analyzed for a time window between 1:00 PM
and 2:00 PM on a given day, then task 56 associates the
particular radio program with the portion of the survey data
collected between 1:00 PM and 2:00 PM. Preferably, the radio
program is associated chronologically with the portion of
survey data, which may include periodic time stamps for
synchronization purposes. Those skilled in the art will
recognize that task 56 may alternatively be performed during
tasks 48 and 50.
Following task 56, a task 58 is performed to present a
survey report to the user. Specifically, task 58 communicates
an audio representation of at least one radio program to
access site 16 and displays a visual representation of the
survey data at access site 16. Task 58 utilizes conventional
data characterization and presentation techniques known to
those skilled in the art. As a result of the synchronous
association performed by task 56, the temporal relationship
between the audio representation and the visual representation
is substantially equivalent to the temporal relationship
between the radio program and the survey data. Thus, a user
at access site 16 hears information that allows the user to
identify the radio programming while simultaneously viewing
survey results for data collected when the radio programming
was broadcast. In addition, the user preferably may change
operational parameters such as the radio station being heard
or the specific time window.
X189414
The preferred embodiment delays the presentation of
survey results for less than approximately 24 hours. This
time shifting allows the selected radio programming and survey
data to be compressed in time to enable the user to quickly
analyze the desired time window. However, nothing prevents
the present invention from presenting the audio and visual
representations in substantially real time to access site 16.
Such real time processing may utilize a full time modem
connection or other data transmission device between survey
site 16 and central office 12 to facilitate immediate
downloading of survey data.
According to the preferred embodiment, broadcast/survey
association process 42 may be performed for relatively short
time intervals. If the time resolution is reduced to
approximately one minute or less, then the listening audience
reaction to a particular song or commercial may be analyzed.
Because survey system 10 preferably utilizes a multiplicity of
survey sites 14 operating in parallel, a statistically
significant amount of data may be gathered over a short period
of time. Thus, survey system 10 is capable of providing data
related to radio audience listening preferences on a song-by-
song basis.
The specific format of the survey report, which is based
upon the collected survey data, may vary from system to
system. For example, the visual representation may be a
graphical representation of the audience listening percentage
for a particular radio station, the approximate number of
listeners at a given time, a statistical comparison between
two or more radio stations, and the like. Data presentation
device 34 may also be configured to produce a printed survey
report for a specific time period or a long term compilation
of measured data.
Following task 58, broadcast/survey association process
42 ends. At this time, process 42 may allow the user to enter
different parameters or alter the current visual display. Of
course, if process 42 is continuously monitoring radio
11
2189~1~
broadcasts and collecting survey data, then task 58 may exit
into query task 52 to monitor for another survey request. It
will be apparent to those skilled in the art that process 42
may be performed concurrently for a plurality of survey sites
and/or a plurality of access sites.
Referring to FIG. 3, survey data collection procedure 46
is depicted as a flow diagram. As described above, procedure
46 is performed during broadcast/survey association process
42. Procedure 46 utilizes known RF-electronic detection
technology to detect LO signals emitted from tuners installed
in vehicles passing survey sites 14. Procedure 46 may be
repeated for a plurality of different LO frequencies
corresponding to different radio station transmitting
frequencies. Control unit 20 may be programmed with the
specific LO frequencies corresponding to the local radio
stations, or programmed to bypass certain frequencies that are
exceptionally noisy or rarely tuned to. The data may be
accumulated at control unit 20 for further processing or
downloading to central office 12.
To minimize the collection of erroneous data, procedure
46 utilizes an LO detection threshold (described in detail
below) such that relatively strong signals, corresponding to
actual LO signals emitted from tuners, are detected. The LO
detection threshold reduces the corruption of the survey data
that may otherwise be caused by the unequal detection of low
strength LO signals associated with radio stations located in
different parts of the frequency band. Thus, procedure 46 is
preferably performed upon the detection of a signal having a
signal strength greater than the LO detection threshold.
Survey data collection procedure 46 begins with a task
60, which initializes a call record that contains data related
to the received LO signal. The call record may include data
such as the date, the survey site location, the detected LO
frequency, the average signal strength, and a time stamp.
Following task 60, a query task 62 tests whether the LO signal
is still above the threshold strength. Procedure 46 repeats
12
2189414
query task 62 until the LO signal falls below the threshold
strength. when the detected LO signal falls below the
threshold, a task G4 completes the call record by, for
example, adding a final time stamp. After task 64, a task 66
stores the completed call record, preferably in control unit
20 (see FIG. 1). Following task 66, procedure 46 ends.
As described above, a site-specific LO detection
threshold is preferably utilized by each survey site 14. To
establish the threshold values, a calibration process 68 is
performed for each survey site 14. Process 68 is depicted in
FIG. 4. Process 68 may, but need not, involve human
interaction at central office 12. According to the preferred
embodiment, process 68 is initiated by a remote command from
central office 12. Such remote control is desirable when
survey system 10 includes a multiplicity of individual survey
sites 14, which may be widely distributed over a large
geographical area. Those familiar with this technology should
appreciate that manual calibration of survey sites 14 may be
unreasonably time consuming and costly.
Calibration process 68 begins with a task 70, which
obtains noise levels associated with the particular LO
frequencies under analysis. Task 70 records any RF signals
that are detected by receiver 18 during "quiet" times, i.e.,
when no tuner-generated LO signals are being detected at
survey site 14. If an actual LO signal is detected, then task
70 ignores it for calibration purposes. Survey system 10
distinguishes actual LO signals from noise by comparing the
signal strength levels, the variation of the signals over
time, and/or other characteristics. Eventually, task 70
produces a noise level distribution over time for each LO
frequency.
Following task 70, a task 72 is performed to select a
maximum noise level associated with the noisiest one of the LO
frequencies. Due to the operating environment, commercial
radio channel spacing, and other factors, certain LO
frequencies may be noisier than others. In addition, the
13
2189414
noise level for a single frequency may vary with time.
Calibration process 68 may perform task 70 for a continuous
24-hour period (or longer) to ensure that the noisiest LO
frequency is identified at its noisiest time of day. Thus,
task 72 analyzes the noise level distributions for each LO
frequency to select the maximum noise level.
After task 72, a task 74 establishes the LO detection
threshold such that it is greater than the maximum noise level
selected by task 70. The LO detection threshold is
established to ensure that LO signals are detected without
statistically skewing the survey data. For example, if the
threshold is set too low, then survey results may be skewed to
favor noisy LO frequencies at the expense of other LO
frequencies, whether or not actual tuner-generated LO signals
are present. After task 74 is performed, calibration process
68 may be suspended (as indicated by the ellipses) as survey
system 10 performs broadcast/survey association process 42.
Thus, process 68 may be performed while process 42 is carried
out during a normal operating mode.
Eventually, calibration process 68 prompts a query task
76, which tests whether previously collected survey data is
substantially inaccurate. Query task 76 is preferably
performed at central office 12. Query task 76 may be
responsive to a system operator input at central office 12 or
one of access sites 16. For example, the system operator may
analyze raw or formatted survey data or observe a visual
representation of the survey data. Using his or her know-how,
the system operator may decide that the survey data contains
aberrant or unrealistic data. Alternatively, computer 26 may
include fault detection programming that periodically analyzes
previously collected survey data to determine if it is
erroneous or inaccurate. Inaccurate survey data may be
indicated for any single survey site 14 if, for example, the
average length of time that a particular LO frequency is
detected is significantly longer than for other frequencies.
As another example, if the average detected signal strength
14
2189414
for a first frequency is significantly higher than the average
detected signal strength for a second frequency at a given
survey site 14, then the data for either the first or the
second frequency may be inaccurate. If inaccurate data is
detected for a particular frequency over a time period, then
survey system 10 preferably discards all of the data collected
from that particular survey site 14 during the time period to
ensure that the survey results are not corrupted. Query task
76 may repeat while survey system 10 operates in a normal mode
until it determines that previously collected survey data is
inaccurate. When this occurs, a query task 78 is initiated.
Query task 78 determines whether the particular survey
site 14 is to be recalibrated. As with query task 76, query
task 78 may be responsive to either a human or a computer-
generated decision. For example, survey system 10 may be
subject to monthly or other periodic calibrations that are
controlled from central office 12. Alternatively, if highly
inaccurate data is being produced by a particular survey site
14, then recalibration may be necessary. If query task 78
determines that recalibration is necessary, then process 68 is
reentered at task 70. If recalibration is not required, then
a query task 80 may be performed. Query task 80 determines
whether the current LO detection threshold is to be adjusted.
This determination may, for example be based on the
characteristics of the survey results. Alternatively, system
operators may wish to analyze the effects of different
threshold values. Again, query task 80 may be responsive to
either a system operator command or a command from computer
26. If query task 80 determines that the current LO detection
threshold is not to be adjusted, then calibration process 68
ends. Process 68 may exit at this time for further analysis
of the inaccurate data or additional system maintenance.
If query task 80 determines that the current LO detection
threshold is to be adjusted, then a task 82 is performed.
Task 82 remotely adjusts the LO detection threshold from
central office 12. This feature allows system operators to
2189414
interactively calibrate survey sites 14 without physically
having to travel to survey sites 14. After task X32 adjusts
the threshold, calibration process 6B is suspended while
survey system 10 operates normally. Thus, survey system 10 is
preferably configured with an interactive and remotely
operated calibration routine that facilitates the periodic
maintenance of survey sites 14 via central office 12.
In summary, the preferred embodiment of the present
invention provides an improved system for estimating listening
audience preferences in response to radio broadcast
programming. A relatively large quantity of survey data based
upon vehicle radio usage may be collected in large or small
geographical areas, which allows the system to detect radio
audience preferences over relatively short time intervals.
The preferred embodiment provides radio survey data
synchronized in time with radio programs broadcast during the
time period that the survey data is collected. In addition,
the present invention is capable of generating and presenting
survey results a short time after the raw survey data is
collected.
The above description is of a preferred embodiment of the
present invention, and the invention is not limited to the
specific embodiment described and illustrated. For example,
the specific hardware implementation of the described
embodiment may be varied to achieve equivalent results. In
addition, the specific tasks of the operating processes
described herein need not be performed in any particular
order, and the individual procedures are not restricted to
particular operating components. Furthermore, many variations
and modifications will be evident to those skilled in this
art, and such variations and modifications are intended to be
included within the spirit and scope of the invention, as
expressed in the following claims.
16
