Note: Descriptions are shown in the official language in which they were submitted.
CA 02455829 2004-01-21
3914 0007
METHOD AND APPARATUS FOR A SATELLITE POSITIONING-BASED METERING
SYSTEM FOR USE IN TRANSPORT-RELATED APPLICATIONS
Technical Field of the Invention
This invention is related to the fields of Intelligent Transportation Systems
and
Transportation Demand Management, more specifically to applications to user-
pay or usage-
pricing such as parking pricing, road use pricing (also known as congestion
pricing, electronic
toll collection (ETC) or electronic road pricing (ERP)), and pay-as-you-drive
insurance.
The background of the problem: Parking Pricing
Parking is a large industry, especially in urban areas of the developed world.
Enormous sums of money flow through this industry which is straddled with well-
known
inefficiencies and related losses for all stakeholders.
Parking is a significant expense for many urban dwellers in the developed
world.
For some, parking may be the 8th or 10th single largest expense following
mortgage, vehicle,
food, education, clothing, etc. The activity of finding and paying fora
parking spot is a nuisance
at best. The confusion regarding how much to pay and how long to pay for, or
the frustration of
overpaying, underpaying, leaving meetings prematurely or risking/receiving
fines adds to the
stress of using parking services.
The business of collecting parking fees is a complex business for
owner/operators of parking lots. It is subject to large volumes of
bookkeeping, pilferage and
user abuse (estimated in some segments of the industry to be 10-15%), wages
for attendants,
high purchase and maintenance costs for ticket vending machines, lift gates,
kiosks, meters,
pay+display machines, and complex and varied loyalty arrangements with local
businesses
wishing to attract customers to their shops.
- 1 -
CA 02455829 2004-01-21
3914 0007
METHOD AND APPARATUS FOR A SATELLITE POSITIONING-BASED METERING
SYSTEM FOR USE IN TRANSPORT-RELATED APPLICATIONS
Technical Field of the Invention
This invention is related to the fields of Intelligent Transportation Systems
and
Transportation Demand Management, more specifically to applications to user-
pay or usage-
pricing such as parking pricing, road use pricing (also known as congestion
pricing, electronic
toll collection (ETC) or electronic road pricing (ERP)), and pay-as-you-drive
insurance.
The background of the problem: Parking Pricing
Parking is a large industry, especially in urban areas of the developed world.
Enormous sums of money flow through this industry which is straddled with well-
known
inefficiencies and related losses for all stakeholders.
Parking is a significant expense for many urban dwellers in the developed
world.
For some, parking may be the 8th or 10th single largest expense following
mortgage, vehicle,
food, education, clothing, etc. The activity of finding and paying for a
parking spot is a nuisance
at best. The confusion regarding how much to pay and how long to pay for, or
the frustration of
overpaying, underpaying, leaving meetings prematurely or risking/receiving
fines adds to the
stress of using parking services.
The business of collecting parking fees is a complex business for
owner/operators of parking lots. It is subject to large volumes of
bookkeeping, pilferage and
user abuse (estimated in some segments of the industry to be 10-15%), wages
for attendants,
high purchase and maintenance costs for ticket vending machines, lift gates,
kiosks, meters,
pay+display machines, and complex and varied loyalty arrangements with local
businesses
wishing to attract customers to their shops.
- 1 -
CA 02455829 2004-01-21
3914 0007
The business of parking enforcement on the part of municipalities is complex
and
expensive. Meters and pay+display machines clutter municipal streets and
parking lots while
their cost erodes parking revenues. Once costs are posted and meters are
calibrated, changing
prices or price regimes is complex, costly and time consuming, minimizing the
opportunity to
use parking as a traffic demand management (TDM) tool.
In some areas, parking is scarce while many small locations that could be used
for parking are not used because of the problem of uneconomical metering and
enforcement.
Conversely, the shortage of parking in some locales encourages drivers to use
such parking
spaces without payment ("spillover"), sometimes using spaces intended for the
use of retail
customers, causing business losses to retail owners.
This invention addresses all of these issues.
A number of patents have been issued for parking and road pricing systems
(Table 1). This invention addresses some of the disadvantages inherent in such
systems.
Table 1 Patent Documents
Jurisdiction and Serial No. Issue or Filing Date
U.S. Patent No. 5,490,079 February 6, 1996
U.S. Patent No. 6,493676 Issued December 10, 2002
U.S. Patent No. 5,721,678 Issued February 24, 1998
U.S. Patent Application No. 20030146852 Filed August 7, 2003
PCT Patent Application No. PCT/DE00/02487 Filed July 25, 2001
PCT Patent Application No. PCT/DE01/03947 Filed October 16, 200'1
The background of the problem: Congestion Pricing/Road Pricing
The practice of charging tolls for road use is well established in many
countries.
The use of toll booths to gate the entrances and exits to highways, bridges
and tunnels whether
manned to collect cash or gantries to read RF1D/DSRC devices are well
understood. Concepts
of high occupancy toll (HOT) lanes are also becoming more familiar. Several
cities already
employ methods to charge tolls for whole downtown districts (e.g. London and
Singapore) using
a variety of technologies such as pre-purchased passes (displayed in the
window of the
vehicle), gantries for RFID readers, license-plate recognition, etc. All of
these systems require
considerable physical infrastructure and considerable labour for enforcement,
both of which are
costly and inefficient both operationally and from an economic pricing
perspective.
- 2 -
CA 02455829 2004-01-21
3914 0007
In addition to the issue of construction and operating costs, the critical
issues of
responsiveness and flexibility need to be addressed. As much as a system
relies on physical
infrastructure (booths, gantries, stationary cameras) that system is
inflexible, costly to modify
and rigid in its ability to address changing needs. The need for on-board
meters that can
determine vehicle position and communicate wirelessly with a central service
is increasingly
apparent and in demand. The need to price any road within a region in order to
manage traffic
and congestion implies that such means are the only ones tenable, since the
infrastructure
costs for cameras and gantries on every roadway make them inefficient in such
applications.
The present invention addresses all these road pricing issues.
The background of the problem: Pay-As-You-Drive (PAYD) Insurance
Insurance premium calculations rely on risks as they are understood to be
affected principally by gender, age, value of vehicle, history of vehicle
model, and driving record.
The risk contributions from issues such as where one drives, when one drives,
or how much
one drives (or where one parks) are not included in a reliable or equitable
manner and remain
as cost externalities. For this reason half of motorists pay too little in
premiums while half
overpay. To resolve this inequity and to bring efficient market principles to
bear on the pricing of
insurance premiums for motorists, the concept of PAYD has been developed
within the
insurance industry. Technology trials to date have included monitoring speed
and distance
using positioning technology and using odometer readings to help assess
premiums.
Technology dedicated to PAYD implies that insurance premiums would have to be
incremented
accordingly to pay for that technology.
The present invention addresses all these insurance premium pricing issues.
Background: Efficient Pricing for Parking and Congestion
This invention specifically sets out to resolve the problem of efficient
pricing for
parking and road use guided by the application of any or all of =the following
principles many of
which are from William Vickrey (Vickrey, W.S. (1969). "Congestion Theory and
Transport
Investment". American Economic Review (Papers and Proceedings) 59, 251-261).
Pertaining to both parking and congestion:
(1) Rates can be set to vary according to location and time of day in an
arbitrarily
timely, detailed, and continuous manner;
- 3 -
CA 02455829 2004-01-21
3914 0007
(2) Pricing may be graduated throughout the period of use, as opposed to being
fixed for a particular time, a particular starting point or a particular
ending point;
(3) Usage rates may be changed without modification of physical signage,
without modification of physical meters and without modification of in-vehicle
devices;
(4) Usage rates may vary smoothly over location and time (no abrupt shoulders
in rate curves);
(5) Pricing can be calculated by the minute.
Specific to parking pricing:
(6) Additional, graduated, parking charges can be levied in lieu of parking
tickets;
(7) The management of and collection for parking usage should be relatively
free
of theft and vandalism;
(8) Charging for small lots with only a few or even only one parking spot can
be
economically viable.
Specific to road pricing or congestion pricing:
(9) To manage congestion most efficiently, charges should be influenced by the
specific impact a particular trip has had on the congestion conditions at the
time of the trip,
rather than be based on a schedule fixed in advance;
(10) Congestion Pricing can be calculated over arbitrarily small trip segments
rather than from one observation point to another in order to allow the
maximum flexibility to
control congestion;
(11) In order to assure fair pricing of service vehicles such as taxis, trip
charges
can be calculated and made available in realtime.
All of the above principles, excepting 6, 7, and 8 apply to insurance pricing
as
well as they do to road pricing.
The Technology Background
Definitions: In this document, a continuous log of a vehicle's position is
gathered
and retained ("position-log"). A position-log is a collection of either
parking episodes ("parking-
log") or journey segments ("journey-log"), or both. journey segments are
further divided into the
time-marked path of travel ("track-log") and the correlated time-marked
congestion conditions
("congestion-log").
- 4
CA 02455829 2004-01-21
3914 0007
With respect to parking pricing, this invention pertains to the ability to
measure
the position, arrival time and departure time ("parking episode") of a parked
vehicle with
sufficient accuracy on which to base a reliable and trustworthy billing
system. A parking
application that uses this technology typically involves three parties: a
vehicle operator (or its
owner/registrant); an owner/operator of a parking lot/spot, and a metering and
billing service or
agency that collects fees from the vehicle owner on behalf of the lot owner. A
vehicle owner
registers a vehicle in a parking agency-service and attaches the subject
location apparatus to
said vehicle. A parking lot/spot owner/operator subscribes to a metering and
billing service and
permits registered vehicles to park on his lot/spot without immediate pre- or
post-payment or
any other form of parking metering. The parking agency-service detects the
exact parking
position (spot), time of arrival and time of departure and calculates a bill
accordingly. In many
cases (sufficiently open sky), there is no requirement for ground-based
equipment, excepting
the subject apparatus, which is installed on or in the vehicle. In some cases
(e.g., underground
parking) it may be necessary to integrate with other technology such as RFID
to open gates and
to confirm arrival and departure times.
To utilize this invention and its related service for parking pricing, a lot
owner/operator should provide an accurate spatial description in the form of
bounding polygons
of the parking lot/spots to be billed and a set of parking rules (times,
rates, free initial period,
overtime penalties, etc).
With respect to road pricing, this invention pertains to the ability to
measure the
track-log and congestion-log of a vehicle with sufficient accuracy on which to
base a reliable and
trustworthy billing system. Its application typically involves three parties:
a vehicle operator (or
its owner/registrant); a road authority, empowered to collect usage tolls on
selected or all
roadways, and a metering and billing service that collects from the vehicle
owner on behalf of
the road authority. A vehicle owner registers a vehicle in a road-use
subscription service and
attaches the subject apparatus to said vehicle. A road authority operates or
subscribes to a
metering and billing service and permits registered vehicles to drive on its
roadways without
immediate pre- or post-payment or any other form of metering or collecting.
This invention
detects the exact route (location and time) and calculates a bill accordingly
and charging or
debiting the owner's account in an automated fashion. A major advantage of
this approach is
that gantries to hold RFID/DSRC or LPR (license plate recognition) technology
are unnecessary
for metering, although LPR, strategically mobile, likely has a critical role
in enforcement.
- 5 -
CA 02455829 2004-01-21
3914 0007
To utilize this invention and its related service for road pricing, a road
authority
must provide an accurate spatial description in the form of driving lanes,
lane usage rules, a
schedule of fees or rules to derive fees in the contexts of location, time-of-
day, and/or relative
congestion.
With respect to pay-as-you-drive ("PAYD") insurance this invention is applied
in a
similar manner to that of road pricing except that premium rates would be set
based on risk
calculations. Hence the digital maps would be essentially the same, but the
premium
calculations would differ from road pricing calculations.
In any of the cases of parking pricing, road pricing or insurance pricing,
business
issues such as billing, collecting, and proof of subscription are basically
well known. Issues of
technology-enabled compliance monitoring may be handled by license-plate
recognition
technology.
In this document, "GPS" is used broadly to refer to any reliable positioning
system, civilian or military, GPS, GNSS, Galileo, GLONASS, LORAN or other. In
North America
the positioning system most likely to be used would be the US system
specifically known as
GPS, but this invention can use signals from any instance of a reliable
positioning system,
including those not space-borne.
Space-borne positioning systems typically have three major components: Space
Segment (the satellites), Ground Segment (the measurement, calibration and
communication
systems that keep satellites on track and accurate), and the User Segment (the
end user
devices that are required to apply the GPS signals. The methods and process
described here
are incorporated within the User Segment.
There are a large number of patents granted or filed that address the
question:
"Where am l?" These describe devices, methods and components to utilize GPS
data in civilian
contexts to locate people, vehicles or assets, to deliver context and location
specific information
to a user of a GPS-enabled device or to navigate, optimize, manage, rescue, or
track. Such
information may be related to distance estimation, trip planning, realtime
navigation, surveying
applications, playing golf, location-dependent advertising, location-specific
advice or
information, and many other applications. Many of these patents address
receiver design,
accuracy, filtering (noise removal), miniaturization, etc - i.e. they address
reliability, accuracy
and dependability of signal reception. Similarity, there are many patents to
address the E-911
initiative for locating cell phones (many E911-related location methods do not
currently utilize
- 6 -
CA 02455829 2004-01-21
3914 0007
GPS). Inexpensive receivers, based on these inventions boast single
measurement accuracies
in the range of 3-10 meters.
Another subset of User Segment inventions address the question "Where is the
vehicle, asset, child, parolee, or patient?" on behalf of an owner, parent,
parole officer or care-
giver. These tracking, monitoring, optimizing or security applications
generally utilize a display-
less receiver-transponder that determines and transmits its location to a
service that can display
or interpret the location of the device in a central service environment. Some
transmit in real
time, while others store successive locations ("tracklog", or "breadcrumbs")
for later use or on-
demand upload to a service or a PC. The method and apparatus described in the
present
invention belongs to this subset of location applications.
Characteristics of this subset of inventions include a small, dependable,
receiver/transponder that is often accompanied by a dedicated processor and
configured for a
specific purpose, a service component (data center), and a telecommunications
link and
protocol.
With regard to devices that capture and store tracklogs there are a number of
properties and constraints that limit prior art in its application to the
problem of metering
vehicles. Limitations of prior art that are addressed by this invention
include:
(1) Accuracy. Parking requires reliable, inexpensive, meter or sub-meter
accuracy. Accuracy requirements are similar for road pricing, when
differentiation between an
HOT lane and an adjacent, non-HOT lane is required. This currently requires a
more expensive
GPS device to gather sufficiently accurate positioning signals. This invention
includes: (a) on-
board filtering and compression 6, 10 specific to the properties of parking-
logs, (b) a specialized
journey-log compression algorithm 12 that conserves critical congestion
information that is lost
in prior art for tracklog compression, and (c) post-processing 28 specific to
the problem of lane-
of-travel identification for multiple users for the road pricing problem.
(2) Storage. It may be necessary to store position-logs for days or months to
conserve telecommunication bandwidth, to utilize and optimize DSRC upload
technologies, or to
wait until a vehicle is powered on to provide power for transmission. This
invention includes a
compression technique specific to the problem of parking 10, and a second
compression
technique specific to the problem of congestion 12. In addition, on board
memory and memory
management algorithms 16 manage store-and-forward processes for these
compressed
position-logs.
- 7 -
CA 02455829 2011-08-19
(3) Service viability. Redundancy is required to ensure uninterrupted metering
and billing. This invention includes a method to use two peer devices to
provide mutual failover
and failure notification to the service component.
(4) Power consumption. Most such devices either require frequent battery
replacements or take power from a vehicle. To make installation easy and non-
invasive, either
battery or solar power may be preferred. Although power source is not a
consideration, here,
this invention includes the design of collection, compression and uploading of
data in a way to
conserve scarce power and telecommunications bandwidth. In addition, device
hibernation is
used during parking episodes once an accurate position is locked down.
(5) Telecommunication consumption. To be always-connected (critical only in
certain applications) communication costs can become prohibitive. Until this
changes, methods
are required to store data on-board until most effectively uploadable. This
invention includes
compression, store-and-forward algorithms, short-range wireless uploads (e.g.
DSRC or UWB)
to pre-empt GPRS or equivalent uploads, and optional learning algorithms to
reduce
communication costs. In this way it is possible to upload off peak as well as
in a daily or weekly
cycle as best suits the application 18.
SUMMARY OF THE INVENTION
In one aspect, the present invention broadly consists in a system for
monitoring,
measuring,or usage metering of a vehicle comprising: an apparatus mounted on
or in the
vehicle, comprising:
= a receiver for receiving positioning signals from a signal source remote
of the vehicle;
= a processor for forming a continuous, time-marked position-log from the
positioning
signals;
= a storage element for storing the position-logs; and
= a wireless telecommunication element, such as GPRS, to transmit position-
logs to a
remote processing system;
= a central processing system in communication with the apparatus
comprising: a central
telecommunication element to receive position-logs from the apparatus;
= a central storage element for storing digital maps and databases for
containing road
usage fees, premium rules, parking fees and schedules; and
= a central processor to calculate user fees based on use of the vehicle.
- 8 -
CA 02455829 2011-08-19
In another aspect, the present invention broadly consists in a method for
monitoring, measuring, or usage metering of a vehicle comprising the following
steps:
= using a mobile communication device mounted on or in a vehicle to
performs steps
comprising the following: receiving signals for generating a position-log for
a vehicle
from a remote signal source; determining for each element of the position-log
whether
it is part of a journey-log or a parking episode; marking the beginning and
end of each
journey-log segment and parking episode; applying distinct compression methods
for
journey-log segments and parking episodes; retaining compressed congestion
information within the journey-log segments; storing position-logs; sending
the position-
logs to a central processing system on a scheduled or demand basis; and
= using the central processing system to perform steps comprising the
following:
receiving transmissions of position logs from vehicles; processing the
received
position-logs for final error removal and data collation; for each parking
episode,
establishing whether it was located in a subscribing parking spot and
deterrnining the
parking fee payable; for each journey-log, determining any road-usage fee due
or
insurance premium due.
DETAILED DESCRIPTION
The key issues that this invention addresses are cost, accuracy, reliability,
flexibility and multiple concurrent purposes. The key elements used to address
these enables
the use of inexpensive receivers to gather accurate measurements that are
inexpensive to
transmit and can be deployed in a multi-purpose system that is unlikely to
fail.
With regard to the areas of parking pricing, road pricing, or insurance
pricing, this
invention is distinct in one or more of the following ways:
(1) The subject invention handles road pricing, parking pricing, and insurance-
pricing equally effectively, and possibly concurrently.
(2) Because satellite positioning technology is used as the preferred location
mechanism, there may not be a requirement for external, ground-based
equipment, especially
such systems utilizing gantries for DSRC or LPR in order to meter compliant
vehicles. Although
it is reasonable to consider the use of mobile or even gantry-mounted LPR as
enforcement
tools, that is a deployment decision independent of metering and this
invention.
- 9 -
CA 02455829 2011-08-19
(3) There is no requirement for a user interface for the on-board component.
In
fact a user interface may make it unnecessarily complex while diminishing
system reliability and
threatening system viability.
(4) There is no requirement for on-board information regarding the
application. In
particular, there is no on-board need for digital maps, payment systems, or
usage rules and fee
schedules for parking, road use, or insurance. This provides for maximum
flexibility by allowing
for the addition or deletion of roads or parking spaces and the modification
of times or prices
without download to the on-board device. Compared to prior art for transport
pricing using
positioning technology, this invention provides for an extremely thin on-board
environment.
(5) Because of the nature of the security algorithms used, there is a greatly
enhanced level of security and privacy protection, thereby alleviating
personal privacy concerns.
Figure 1 is a depiction of the major components in relationship to each other
of
this preferred embodiment. One or more apparatus 102 mounted in or on Vehicle
104 receives
position signals from satellites 106, and after some processing forwards
compressed,
intermediate results to central processing system 108, for further processing,
generation of
invoices for parking, road-use and insurance, and for the generation of maps,
and various data
feeds.
Figure 2 is a flowchart which illustrates an overview of the methodology of a
preferred embodiment of the system including an on-board apparatus 102 and
data center
(also known as central processing system) 108 components and all parking
pricing and road-
pricing components. Not all elements are required for every correctly
operating instance of the
system.
Figure 3 shows 100 consecutive data points (position readings) 1110 taken from
a
stationary platform for a preferred embodiment; the circle 112 is radius 5m
around the actual
platform location. This conforms to the specification that 95% of the position
data points fall
within 5m (for the particular device used). Theoretically, as a greater and
greater sample size is
accumulated the average tends to the true location while the standard error
tends to zero. In
practice, a bias may be present, and there are published methods to remove
many of those, as
well.
Figure 4 depicts a parking situation for a preferred embodiment in which two
different subscribers (e.g., a municipality and a private lot operator) have
spots within three
meters of each other. Note that the targeted vehicle 104, in a spot in the
private lot, records
- 10 -
CA 02455829 2004-01-21
3914 0007
many individual readings that are not within the spot it is parked in. Indeed
a few stray readings
116 are within or even beyond spots in the municipal area. This invention
includes processes to
resolve this problem.
Figure 5 depicts an exaggerated and artificially tidy (random variation is not
depicted) illustration of a situation in which cars 104 traveling in an HOT
lane produce journey-
logs 120 whose absolute positions are biased but whose relative positions are
correct. Using
data from a group (mob) of vehicles 122 that are concurrent and co-located,
statistical
techniques allow categorization of cars 104 into lanes of travel and
correction for any bias. This
is intended to be used only in cases of ambiguity or requirement of proof of
charges.
Addressing Cost in a Parking Application
In the case of parking pricing applications, cost is addressed by applying a
process and method for filtering parking-logs to gain sufficient accuracy of
position and time
from repeated measurements 6, allowing the use of inexpensive receivers. Cost
is also
addressed by the use of compression 10, specifically designed for parking-logs
in order to
reduce telecommunication expense.
On-board, filtering for compression and accuracy of parking information relies
on
properties of GPS data, the principle property being that each model of
receiver has known
statistical error behaviour - for example, many inexpensive devices utilizing
DGPS specify that
95% of the measures taken are within three to five meters of the true position
(see Figures 3
and 4). Such statistical properties can be measured and highly reliable
estimators can be
derived regarding the number of successive measurements that are required to
be certain that a
vehicle's position data exhibits stationarity (that quality of a process in
which statistical moments
(e.g., mean and covariance matrix) of the process do not change with time) and
how many
samples must be accumulated to state a vehicle's position within an arbitrary
accuracy. As an
example, these statistical properties allow a statement such as: "Given a
sampling rate of one
per second and a device accuracy of 3 m, the position of a stationary vehicle
can be determined
to within 1 meter in M minutes within a statistical error of 10-6. This
ability to generate a
sufficiently accurate position with an inexpensive GPS receiver is key to this
invention relative to
parking. Once an accurate position is assured, data collected for the settling
period can be
reviewed for inclusion in the "parked" time, so that a majority portion of the
M-minute settling
period can be allocated for an accurate measure of the start time. The same
procedure is used
in reverse for the finish time, to enable parking pricing to the nearest
minute.
- 11 -
CA 02455829 2004-01-21
3914 0007
This filtering process proceeds (within the apparatus mounted in the vehicle)
as
in the following example (the control parameters for this algorithm are not
fixed, and the ones
used here are for illustrative purposes only):
Each minute, collect 60 samples with an inexpensive location device.
For each pair of successive, 1-minute samples, calculate the statistical
moments
required to determine statistical stationarity and statistical equivalence
between two successive
minutes. Two successive minutes that satisfy these criteria become a candidate
for the
initiation of a parking episode, while the pooled mean becomes the candidate
position of a
parked car.
Continue this during minutes{3,4...M} to confirm within pre-set accuracy
levels
that the vehicle's location measurement exhibits stationarity at position P
and the time of minute
M; else discard this as a candidate parking episode.
Recalculate in reverse through the data from minutes {M-1, ... 2,1} to assert
an
arrival time as close as possible to the actual arrival time.
Continue sampling and testing the statistical properties of the location data
for
each minute {M+1, M+2,...} against the hypothesis that the vehicle's location
measurement
continues to exhibit adequate stationarity to confirm a sufficiently exact
location.
Alternatively, given that the incorporated motion detector has registered a
motionless state and the stationarity of the positioning signal allows
determination of an precise
(but possibly biased) position, the device may hibernate to be awoken only
upon the detection
of motion.
When the hypothesis fails, or the motion detector detects motion, establish
the
conclusion time and summarize the entire event as a parking episode (position,
start time, end
time).
This description is the simplified core of an operational algorithm to
accurately
identify and measure a parking episode using an inexpensive receiver. The fact
that GPS
satellites are not geostationary and that a vehicle may be in an urban canyon
(restricts the
number of visible satellites) or beneath foliage (scatter) means that signal
stationarity is not
guaranteed. Hence urban canyon and foliage, as may be resolved, will require
somewhat more
complex statistical analyses, but with the same result: position, arrival and
departure times of
sufficient accuracy on which to build a reliable and trustworthy billing
system.
- 12-
CA 02455829 2004-01-21
3914 0007
This process is not required to determine whether a vehicle is standing still -
that
can be done more effectively with the motion sensor 4 that is incorporated
into this invention.
This process is used to determine whether the vehicle is in a spot for which
it is possible to
receive a statistically stationary reckoning of position. Since this is not
possible in every open-
air, ungated parking spot, all such spots must be pre-measured to determine
suitability for this
method. Currently a large portion of parking spots satisfy this criterion. As
positioning
technology improves - or with the use of more expensive receivers - the
portion of spots that
are eligible will increase. A process for pre-measuring parking spots includes
a survey activity
for each contiguous set of spots (parking lot, or strip of on-street spots).
This survey would
determine whether the desired accuracy can be achieved using a GPS device with
the accuracy
of the receiver used in the subject invention throughout a 24-hour period and
throughout all
seasons (foliage).
There is the special case of a parked vehicle parked in radio shadow
(underground, under foliage, or in an urban canyon). In many instances, noting
that a vehicle
ceases receiving reliable positioning signals information at a certain
position then restarts later
from that same position (as might occur on entering or departing a parking
garage) it is
reasonable to infer a parking episode. The issue of establishing whether the
position inferred is
within a subscribing parking lot/spot will be determined by the central
processing center.
Naturally, in the circumstance of gated parking lots, access to a spot might
be provided via
RFID which provides for existing technology to identify vehicle (customer),
position and time of
parking episode. For this reason links to, and incorporation of, RFID
technology is part of this
invention 4.
Cost, as it pertains to storage and telecommunications usage, is addressed by
including a system and method for compressing position-logs to reduce data
volumes.
In this invention, compression with respect to parking takes advantage of the
fact
that one is only concerned with the time the vehicle is stationary. Hence, the
output of
processing will be position, start time, and finish time. Most data judged to
come from a moving
vehicle or from a vehicle that is remaining stationary relative to the prior
time period may be
discarded in the case of a parking-only application. A few critical measures
of settling times,
including some of the raw or partially processed data, may be retained to
provide additional
flexibility for the central billing process to prove accuracy of determination
of settling time. This
flexibility might be needed to make the necessary adjustments to ensure
unquestionable
fairness of the amount billed, which would now likely be measured to the
nearest minute. As
- 13-
CA 02455829 2004-01-21
3914 0007
well, evidence of the data collected just prior and just after the time of
parking lock-in may be
useful as evidence of the original measurements to prove fairness of the bill.
In general, for a
device that is collecting data every second for 24 hours, the amount of data
to be uploaded to
represent, say, 5 parking episodes would be on the order of 10kb or less
whereas the full raw
data set would be on the order of 1Mb, a compression ratio of 100:1. To this
reduction in
retained data, known compression techniques may be added prior to data
encryption. Principal
among these are well-known delta-encoding and Lempel-Ziv (LZ) techniques,
which taken
together provide an additional order of magnitude of compression, depending on
the data
retained. If all data were discarded excepting position, start and end, one
would expect a
compression ratio of approximately four orders of magnitude compared to raw,
second-by-
second position-logs. This implies that the more process intelligence is
applied to the device,
the less demand one would expect on storage or telecommunication services.
Reliability, as it pertains to assurance that the device has a reduced
likelihood of
being excluded from service by accident, malice or failure is addressed by
using two,
independent, self sufficient, intercommunicating, receive-filter-compress-
store-forward devices
on a single vehicle.
These two devices, identical in everything excepting an IP address, comprise
the
complete on-board component for this invention. This is a two-element, peer-to-
peer net
providing redundancy and mutual fail-over. While only one of the devices needs
to upload its
data to a central processing center, encoded in that upload is information
regarding the health of
both devices. Each peer device would be programmed to trigger an upload as
soon as the
other component failed. System faults can thus be recognized early and
remedied sooner
reducing the likelihood of excluding a vehicle from the network.
In the case of road pricing applications, companion algorithms running at the
central processing center and including the use of concurrent and co-located
journey-logs from
other vehicles can be used to further process a specific journey-log to
identify which lane (in a
multilane highway) a moving vehicle is in 28. This is useful to remove
position bias in order to
distinguish HOT lanes from non-HOT lanes (Figure 5). In addition, post-
processing at the
center can also include additional information to address bias effects as
given by ionospheric
disturbances, etc.
- 14 -
CA 02455829 2004-01-21
3914 0007
Addressing Cost in a Road Pricing Application
In order to enable fully efficient road pricing programs, sufficient
positional
information to determine exactly which lane a vehicle is traveling in must be
retained. This
extreme scenario handles the case of differential pricing of two adjacent,
fully communicating
lanes carrying traffic in the same direction, one HOT the other not (Figure
5). Methods to
optimally compress the position elements of such a journey-log are already
known (for example
Douglas-Peucker or "DP"). These algorithms turn an over-sampled series of
positions into a
smaller series of vectors (start-point, end-point) so that the sarne track is
represented with fewer
data points and with bounded spatial error (the "hull" in the case of DP line
compression). The
present invention makes two modifications 12 to this known compression method
as applied in
preferred embodiments.
The first modification of the DP method is that the event time for the start
of each
compressed data vector is retained in order to establish the exact times of
each new sub-
segment of the track-log. This time-marked track-log allows a fully detailed
breakdown of a
journey for charging by time-of-day and road traveled to provide for any level
of congestion
management according to Principle 1, 2, 4 and 10, above.
The second modification to DP for the creation of a compressed journey-log
that
is incorporated in this invention is the identification, storage, and
compression of congestion
information. One definition of congestion of a system of roads and
intersections is derived from
the likelihood that cars queued at red signals within the system do not clear
those intersections
during a full cycle (red-green-back to red). Naturally, as cars may remain
queued through two
or more light cycles, congestion may back up to a prior intersection causing
gridlock, an
extreme condition of congestion, similar to that caused by a crash. It is
possible to infer and
measure congestion from the velocity changes in a journey-log. A concurrent
collection of
journey-logs would provide invaluable information about specific, temporary
congestion for the
purpose of aiding navigation algorithms used in the kind of automotive
navigation aids presently
in popular use. This would work for normal rush-hour congestion as well as for
congestion due
to crashes, special events, construction, etc.
Additional Impacts Of The Use Of This invention
In the subject invention, congestion information is compressed and inserted on
a
segment-by-segment (vector-by-vector) basis into time-marked, DP output. This
is computed
by taking a second pass through the raw data to extract velocity information.
The result of this,
- 15 -
CA 02455829 2004-01-21
3914 0007
a congestion-log, is retained for each vector (straight segment) of the
compressed DP track
segments. Taken together, this compressed data set of where, when, velocities
and velocity
changes constitutes a complete journey-log, which can be used to satisfy
Principle 9. In its
most sophisticated form, the journey-logs of a mob of vehicles within a
defined traffic area can
be used to back calculate the effect of a single journey on the then-current
congestion
circumstances to derive the optimally efficient congestion charge. Admittedly,
this may be too
complex a concept for realistic vehicle-by-vehicle charge calculations, but
this data is invaluable
to general realtime navigation, optimization of a large array of traffic
signals and certain
components of traffic demand management (TDM) planning. For that reason
collecting,
aggregating and distributing this information is useful even if not used to
calculate fully efficient
congestion pricing for individual trips.
Finally, this congestion information is useful in a multilane (HOT and non-
HOT)
environment (see Figure 5). In the case that raw location information is
insufficient to determine
the lane of travel (due to scatter or multipath error), a group of
simultaneous and co-located trip
segments would show relative positions in a way that allows distinction among
the absolute lane
positions. In the circumstance of an HOT lane that is consistently moving
faster than the
adjacent non-HOT lane, that information will be apparent by comparing the
congestion-logs of
adjacent lanes. Knowing relative information about individual vehicles in a
mob of vehicles
allows the inference of absolute position (lane) of those individual vehicles.
This invention benefits vehicle owners by removing all of the annoyances of
using mechanically metered parking or any form of active toll payment for road
use. It may
benefit owners and operators of parking facilities by reducing equipment
costs, abuses by
vehicle owners and employees and by enabling small, otherwise uneconomical
lots to become a
viable source of revenue. It may benefit road authorities by enabling
economically efficient
electronic road pricing (ERP) to be deployed in the most cost effective
manner, as well as
building congestion databases to be used for planning and control.
This invention implements a wide variety of parking and road-use policy and
allows the flexibility to manage the most efficient (and fair) methods of
pricing these
commodities in order to include both internal and external transport costs,
thereby removing
possible pricing inequities from transport systems.
It will be appreciated that the above description relates to the preferred
embodiments by way of example only. Many variations on the system and method
for
- 16 -
CA 02455829 2004-01-21
3914 0007
delivering the invention will be clear to those knowledgeable in the field,
and such variations are
within the scope of the invention as described and claimed, whether or not
expressly described.
- 17-
