HISTORICAL OVERVIEW OF AUTOMOBILE NAVIGATION TECHNOLOGY
`
`Robert L. French
`Consultant
`3550 Rulen Street
`Fort Worth, Texas 76107
`
`Key developments relating to vehicle
`navigation are listed in Table 1, and are
`discussed by category below. Virtually
`all high-technology automobile navigation
`systems use on-board computers to inte-
`grate and automate two or more of these
`technologies to provide vehicle location,
`heading, routing, or step-by-step guid-
`ance. Subsequent sections describe system
`concepts and functions, and give examples
`of both early and
`emerging automobile
`navigation systems.
`
`Table I. Vehicle Navigation Developments
`
`DATE
`
`TECHNOLOGY
`
`<60 AD
`200-300
`1100-1200
`1906
`1910
`1940
`1964
`1966
`1971
`
`Odometer
`Differential Odometer
`Magnetic Compass
`Gyrocompass
`Programmed Routes
`Loran Positioning
`Satellite Positioning
`Proximity Beacon
`Map Matching
`
`ABSTRACT
`
`Sophisticated automobile navi-
`gation systems are becoming feasible
`largely because of low-cost computer
`technology.
`Computer technology is
`used to add
`artificial intelligence
`to dead reckoning
`navigation tech-
`niques that have been around for
`as
`long as 2000 years, and to automate
`radio frequency navigation concepts
`dating back to WW-11.
`This paper
`outlines the history of
`navigation
`technologies applicable to automo-
`biles, and discusses their integra-
`tion in various systems approaches
`ranging from the mechanical
`route
`guides of the
`early 1900's to the
`high-technology systems now beginning
`to appear.
`
`INTRODUCTION
`
`Navigation is an applied science that
`uses a variety of techniques and proce-
`dures to determine present position,
`heading, and/or direction and distance to
`a destination. For
`centuries, navigation
`technology centered on the use of celes-
`tial observations and compass readings to
`fix the position
`and set the course of
`ships at sea.
`Similar techniques were
`adapted for aircraft navigation, and were
`soon joined by
`radio direction finding,
`triangulation, inertial guidance, and
`satellite positioning systems.
`However, except for a brief period of
`activity starting around 1910, automobile
`navigation had received little attention
`compared to sea, air and space navigation
`prior to the recent flurry of engineering
`projects, concept car showings, and media
`reports. Recent
`developments include a
`pioneering map matching
`system that is
`already on the market in California, as
`well as Ford and
`Chrysler concept cars
`with satellite navigation. General Motors
`has recently launched institutional
`ad-
`vertising foretelling "electronic naviga-
`tion systems that tell you where you are
`and how to get where you're going." In
`addition, there are international develop-
`ments that, in some ways, outpace the
`domestic programs.
`
`CH2308-5/86/0000-0350 $1.00 Q 1986 IEEE
`
`DEAD RECKONING TECHNIQUES
`Dead reckoning is the process of
`determining a vehicle's location and head-
`tng relative to an initial position by
`integrating measured increments and direc-
`tions of travel. A vehicle's
`current
`position in rectangular (x,y) coordinates
`is given by
`+ Icos+(t)dt
`x = x.
`y = yo + jsin+(L)da
`( 2 )
`coordinates,
`where xo, y are the initial
`dL is the %ravel increment, and +(L) is
`the vehicle heading associated with the
`distance increment.
`Most dead reckoning technologies used
`in automobile navigation systems were
`developed long before the
`automobile
`itself. These include the odometer, the
`magnetic
`and the
`differential odometer
`350
`
`(1)
`
`I
`
`IPR2017-02022
`Unified EX1013 Page 1
`
`

`

`The gyrocompass, developed in
`compass.
`in
`1906, has seen only limited usage
`automobile systems.
`Odometer
`Descriptions of odometers started
`appearing in both
`western and Chinese
`literature approximately 2000 years ago
`(1). One early model described by Hero of
`Alexandria recorded distance travelled by
`dropping a
`stone into a receptacle at
`periodic intervals. An early
`Chinese
`odometer struck a drum at distance inter-
`vals of approximately 500 yards, and rang
`a bell at every 10th interval.
`The odometer was the basis for a wave
`of mechanical road guides beginning around
`1910 as a proliferation of automobiles
`created a demand for routing information
`before adequate road signs and maps were
`generally available. Among the
`first was
`the Jones "Live Map" ( 2 ) . This mechanical
`road guide consisted of a turntable driven
`by a gear train and flexible shaft con-
`nected to one of the vehicle wheels.
`Individual routes were "programmed" on
`paper discs with a scale of miles printed
`around their perimeter. The
`discs were
`mounted on the turntable beneath a glass
`cover with a fixed
`pointer, and printed
`road
`directions
`keyed
`to
`specific
`distances from the beginning of the route
`came into view under the pointer at the
`time for execution.
`One of the most sophisticated mechan-
`ical route guides of the era was the
`Routes were
`Chadwick Road Guide ( 3 ) .
`programmed by holes punched in a metal
`disk, and signal arms and a bell
`were
`activated by the punched holes
`as each
`maneuver point was approached. One of ten
`signal arms bearing color-coded
`symbols
`indicating the action to be taken would
`appear behind a window and the bell would
`sound to attract the driver's attention.
`Conventional mechanical odometers for
`automobiles are usually driven by flexible
`shafts attached to the drive train, and
`display distances to the
`nearest 0.1
`mile. Electronic
`odometers, which can
`measure trave.1 increments as small as one
`inch, are often used in automobile naviga-
`tion systems. These provide
`sensor sig-
`nals from a rotating shaft or wheel and
`apply a
`conversion factor to obtain
`dlstance travelled.
`Differential Odometer
`A differential odometer is essen-
`tially a pair of odometers, one each for
`wheels paired on
`opposite sides of the
`vehicle. When the
`vehicle changes head-
`ing, the outer wheel travels further than
`the inner wheel by an amount (AD) that is
`
`change in
`equal to the product of the
`heading (A+) and the vehicle's width (W):
`AD = WA+.
`( 3 )
`Thus, by measuring the differential travel
`of opposite wheels, a vehicle's path and
`heading relative to its starting point may
`be computed using
`algorithms based on
`Equation 3 .
`The differential odometer principle
`was used in a direction-keeping device by
`the Chinese approximately 200 - 300 A.D.,
`perhaps earlier (1). The device was in
`the form of a horse-drawn two-wheel
`cart
`bearing a turntable-mounted statue with an
`outstretched arm. It
`was called the
`"south-pointing carriage" because the
`statue's arm always continued to point in
`its original direction regardless of which
`way the carriage turned as it travelled.
`When changing heading, a gear train driven
`by the
`south-pointing carriage's outer
`wheel automatically engaged and rotated
`the horizontal
`turntable (bearing the
`statue) to offset the vehicle's change in
`heading.
`The differential odometer was tested
`as a basis for automobile navigation by
`Meyer in 1971 ( 4 ) .
`He used the differen-
`tial odometer and a
`mechanical dead-
`reckoning computer to keep track of
`vehicle coordinates and heading.
`Test
`results indicated heading errors averaging
`approximately 20 degrees per mile
`of
`travel. At about the same time, French
`developed and tested an electronic version
`of the differential odometer in an auto-
`matic route guidance system which included
`a map-matching algorithm to maintain much
`higher accuracy (5).
`Magnetic Compass
`The magnetic compass's well-known
`accuracy problems due to anomalies in the
`earth's magnetic field become more severe
`when installed in an automobile because of
`induced fields depending upon vehicle
`heading. In addition, an automobile may
`have a permanent magnetic field of its
`own, and sub-permanent magnetism may be
`acquired or lost when hitting bumps.
`Nonetheless, magnetic compasses have
`long been used as heading indicators for
`automobiles, and modern versions of the
`compass are now
`frequently used as
`a
`component in integrated navigation systems
`for automobiles.
`Whereas the compass
`first appeared around the 11th century in
`the form of a magnetic needle floated on a
`current versions
`liquid surface ( 6 1 ,
`include compact solid state €lux-gate
`compasses with software algorithms for
`compensating errors due to both permanent
`and induced magnetism of the vehicle.
`
`IPR2017-02022
`Unified EX1013 Page 2
`
`

`

`a mag-
`Perhaps the earliest use of
`netic compass in an integrated vehicle
`system was the vehicular odograph, a self-
`contained navigation system for U. S. Army
`vehicles developed during WW-I1
`(7). An
`electro-mechanical system drove a stylus
`to automatically plot vehicle course on a
`map of corresponding scale.
`An odometer
`provided a distance input which was
`resolved into x,y components using servo-
`driven input from
`a photo-electrically-
`read magnetic compass.
`Gyrocompass
`The gyrocompass, which was. paf.ented
`in Germany in 1906 by Auschutz-Kampfe,
`uses the gyroscopic inertia principle to
`maintain a constant reference alignment in
`space. One of the few examples of the
`gyrocompass being incorporated in an
`automobile navigation system is the Lunar
`Roving Vehicle
`(LRV) used on the moon
`where a magnetic compass would be use-
`less. The
`LRV, driven on the moon's
`surface during Apollo Missions 15, 16, and
`17, had a dead reckoning navigation system
`which continuously measured direction and
`distance travelled, and periodically cal-
`culated vehicle position.
`Using gyro-
`compass reference heading
`and magnetic
`odometer inputs, the system could return
`the astronauts to within 100
`yards of
`their origin after a 20-mile journey ( 8 ) .
`MAP MATCHING TECHNIQUES
`Navigation of an automobile from one
`location to another differs considerably
`from sea and air navigation because the
`automobile is essentially constrained to a
`finite network of streets and roads. This
`makes it possible to apply artificial
`intelligence pattern recognition concepts
`to match a vehicle's dead reckoned course
`with a mathematically mapped route
`or
`route network to compensate dead reckoning
`error and to
`determine vehicle location
`with great accuracy.
`Map Mode 1 i ng
`Graph theory is used as a conceptual
`framework for mathematically modeling maps
`of roads and
`streets as internodal vec-
`tors. Each vector
`is the combination of
`distance and direction representing the
`road between two nodes defined
`by their
`coordinates.
`Thus, a particular route
`from a given initial location
`may be
`defined as a unique sequence of vectors.
`The most extensive
`body of
`vector-
`encoded map data
`is the GBF/DIME
`(Geo-
`graphic Base File/Dual Independent Map
`Encoding) System developed
`by the U. S.
`Census Bureau for virtually all Standard
`Metropolitan Statistical Areas
`(SMSA).
`
`GBF/DIME includes street names and address
`ranges between nodes, as well as a variety
`of information
`not relevant to vehicle
`navigation. Although
`a useful resourse,
`this map data base has numerous errors in
`segment connectivity, and represents the
`street network in 1977 (9).
`Since existing map data bases may not
`be used without extensive correction and
`revision, commercial map publishers are
`positioning to
`supply special map data
`(10). One
`bases for vehicular navigation
`automobile navigation systems company,
`Etak, Inc., has undertaken development of
`a special digital map data base €or the
`entire U.S. (11).
`Pattern Matching
`In map matching, the pattern of the
`vehicle's path is analyzed and defined as
`a sequence of vectors deduced from any of
`a variety of dead reckoning processes. As
`the vehicle travels, its measured vector
`sequence is continuously compared with the
`mapped vector sequence. Each time
`a turn
`is executed whose sense, magnitude, and
`location approximate those of a mapped
`turn, the vehicle is presumed to be at the
`mapped location. The matching
`process
`thus removes any dead reckoning
`error
`accumulated since the last turn.
`Vector map matching was demonstrated
`puter analyzed differential odometer sig-
`nals to deduce a delivery vehicle's path
`and match it with programmed routes.
`An
`average location accuracy of 4 feet was
`maintained during extensive testing of the
`map matching process.
`Error Compensation
`In addition to correcting dead rec-
`koning errors, map matching may be used to
`continuously fine tune differential odom-
`eter calibration factors used in computing
`distance increments (12). When an adjust-
`ment is necessary to make the perceived
`location of a turn conform with the pro-
`grammed location, the difference may be
`used to calculate revised calibration
`factors which are weighted
`and combined
`with the prior factors for subsequent use.
`RADIO FREQUENCY TECHNIQUES
`Numerous navigation schemes, includ-
`ing triangulation, phase and pulse trilat-
`eration, Loran, proximity
`beacons, and
`satellite approaches, involve the use of
`radio signals in determining position.
`The later two have received the most at-
`tention as the basis for vehicle location
`and
`navigation systems using
`radio
`signals.
`
`in 1971 by French ( 5 ) . An on-board com-
`
`352
`
`IPR2017-02022
`Unified EX1013 Page 3
`
`

`

`Proximity Beacon
`This approach uses strategically
`located short-range transmitters, and the
`very reception of their location-coded
`signals infers the receiving vehicle's
`instantaneous location, The widest appli-
`cation of proximity technology is for
`automatic vehicle location (AVL) monitor-
`ing systems such as those used for
`monitoring the
`location and status of
`from a central dispatch
`transit buses
`office (13). An on-board system receives
`and stores a location code as the vehicle
`passes a proximity beacon or "electronic
`signpost".
`Upon periodic polling, the
`last beacon location and the distance or
`time since passing the beacon are automat-
`ically radioed to the dispatch computer.
`Several variations of the proximity
`approach, some involving two-way communi-
`cations with equipped vehicles, have been
`investigated for interactive route guid-
`ance (14!. Typically, the driver enters a
`destination code on the vehicle panel for
`automatic transmission to a roadside unit
`as the vehicle approaches instrumented
`intersections. The
`roadside unit, which
`may operate autonomously, or be networked
`with a traffic management system, analyzes
`the destination code and instantly trans-
`mits route instructions for display on the
`vehicle's panel. Alternatively, the road-
`side unit may only transmit its location
`to the vehicle where an on-board computer,
`using stored road network
`data, will
`generate the route instructions from the
`identified location.
`
`Proximity beacon navigation was re-
`searched in the United
`States starting
`with DAIR in the mid-1960's. DAIR (Driver
`Aided Information and Routing System),
`which used roadbed
`arrays of magnets
`arranged in binary code to communicate
`location to passing vehicles, was the sub-
`ject of limited development and testing by
`General Motors (15). It
`was followed by
`ERGS (Electronic Route Guidance System), a
`Federal Highway Administration project of
`the late 1960's
`which promulgated radio
`communication between vehicle and roadside
`units (16). ERGS was discontinued in the
`early 1 9 7 0 ' ~ ~ largely because
`of the
`expensive infrastructure required, but
`similar approaches have subsequently been
`tested in West Germany and Japan.
`
`Satellite
`Satellite positioning has received
`considerable attention as a basis for
`automobile navigation systems. The
`"Transit" satellite system, which was
`implemented by the
`U. S . Navy in
`1964
`(17) , was the basis for the Ford "Concept
`100" car in 1983 (18). Transit receivers
`
`determine location by Doppler analysis of
`signals from a passing satellite.
`The
`satellites are in polar orbit at a height
`of approximately 1,000 kilometers, and are
`longitudinally spaced to give worldwide,
`although intermittent, coverage. Since a
`Transit satellite is not always in range,
`the Ford system included a dead-reckoning
`subsystem (based on odometer and magnetic
`compass inputs) for determining position
`between satellite passes. Selectable CRT
`displays showed vehicle heading, map
`location, etc.
`The Navstar Global Positioning System
`(GPS), which is still in the implementa-
`tion stage with the last of 18 satellites
`to be launched within the next few years
`(17), has been considered as a basis for
`automobile navigation by all of the major
`U.S. motor companies (18, 19, 20). When
`the satellite constellation is complete,
`any point on earth will always be within
`range of at least four Navstar satellites.
`A GPS receiver could determine its
`three position
`coordinates by analyzing
`the travel time of signals from only three
`satellites if the receiver's
`clock was
`precisely synchronized with the atomic
`clocks that time the
`satellite signals.
`However, given the timed signals from four
`satellites, the GPS receiver solves a
`uations for its three
`system of four e
`9
`s (P , Pyl Pz), and for
`position coordinat
`e
`the bias (CB) of
`i
`ts Tess precise quartz
`clock:
`(xl-Px)2+(Y1-Py) 2 +
`(x2-Px) 2 +(y2-Py) 2 +
`(x3-Px) 2 +(y3-Py) 2 +
`(x4-Px) 2 +(y4-Py)2+
`where x,, yn, zn, are position coordinates
`of the nth satellite based on ephemeris
`carried by its signal, and is the
`range
`to the nth satellite.
`Locations may thus be determined with
`an accuracy of approximately 50 feet using
`Navstar GPS P-Code signals intended for
`Department of Defense applications.
`Less
`precise C/A-Code signals intended for
`commercial use will permit location
`determination to within 300-500 feet.
`Although Navstar
`GPS will provide
`continuous coverage when the satellite
`constellation is completed, vehicle loca-
`tion determination may be impaired at
`times due to signal attenuation or reflec-
`tion by buildings, foliage, terrain
`features, etc. Therefore, as in the case
`of Transit, auxiliary dead reckoning is
`required for GPS automobile applications.
`
`353
`
`IPR2017-02022
`Unified EX1013 Page 4
`
`

`

`Other RF Techniques
`Of the various other radio-frequency
`based approaches that have been proposed
`or tested for vehicle location monitoring
`(21), Loran (Long RAnge Navigation) prob-
`ably has the most potential for automobile
`navigation because of its relatively wide
`coverage. Invented
`in 1940 by Loomis,
`Loran is a hyperbolic navigation system
`using multiple pairs
`of ground-based
`master-slave transmitters. U. S. coverage
`is complete along all coasts
`and most
`inland areas except for a band from Texas
`to North Dakota (22).
`a se-
`The master transmitter emits
`quence of electromagnetic pulses which are
`received by a widely separated slave sta-
`on a slightly
`tion and retransmitted
`different frequency. The Loran
`user’s
`receiver picks up both signals and anal-
`yses the difference in arrival time to
`establish a hyperbolic line-of-position
`associated with the time delay. Repeti-
`tion of the
`procedure for a different
`master-slave pair yields the longitude and
`latitude of the receiver at the intersec-
`tion of two hyperbolae.
`The availability of microprocessor-
`controlled receivers with improved per-
`formance at lower cost may make Loran C an
`attractive alternative to proximity beacon
`AVL in some cases (22, 23, 24). But Loran
`has not been pursued extensively as the
`basis for automobile navigation systems,
`perhaps due to incomplete
`geographic
`coverage as well as location inaccuracies
`typically on the order of 600 feet. One
`known use of Loran in a U.S. automobile is
`a General Motors test system which
`also
`includes GPS positioning (19). Loran is
`also used in
`a Nissan delivery vehicle
`navigation system currently being tested
`in Japan ( 2 5 ) .
`SYSTEMS ENGINEERING
`
`Navigation and positioning techniques
`outlined above may be integrated in vari-
`ous combinations for automobile
`systems
`providing a wide range of functions.
`Navigation Functions
`system func-
`Automobile navigation
`tions are broadly categorized according to
`whether they provide
`the driver with
`location information
`only, with other
`information useful in routing selection,
`or with explicit real-time route guidance.
`Location Vehicle location informa-
`tion has several possible forms.
`A
`is to indicate vehicle
`popular concept
`location by the position of a cursor on a
`map display panel. Another concept is to
`display street name and to give position
`
`in terms of address range or distance to
`the next
`cross street. Vehicle
`coordi-
`nates per se are generally not provided.
`Routinq A variety of other informa-
`tion may be useful to the vehicle driver
`in selecting a route to a particular dest-
`ination. Examples include vehicle head-
`ing, direction to destination, distance to
`destination, location of destination on
`map display, highlighted thoroughfares on
`map display, etc. An approach used in a
`few cases displays the route travelled to
`the present location by plotting it on a
`map.
`
`route guidance
`Guidance Real-time
`prompts the driver turn-by-turn over an
`appropriate route to his destination.
`Once the driver specifies destination and
`routing criteria
`(fastest, shortest,
`scenic, etc.) , the system makes all navi-
`gation decisions, freeing the driver to
`concentrate on driving safely. Explicit
`route guidance may
`be in
`the form of
`spoken instructions, displayed symbols
`(e.g., arrows shaped according to the
`maneuver), and/or displayed messages.
`System Concepts_
`of domestic
`Although the wide range
`and foreign systems approaches for auto-
`mobile navigation defies precise categori-
`zation, it is convenient to use the class-
`ifications characterized in Table I1 for
`discussing major system concepts.
`
`Table 11. Automobile Navigation Systems
`
`CLASSIFICATION
`
`TYPICAL CHARACTERISTICS
`
`Dead Reckoning
`Low Vehicular Expense
`Autonomous Operation
`Accuracy Degrades with Travel
`Requires Manual Updates
`Proximity Beacon
`Low to Moderate Vehicular Expense
`Requires Costly Roadway Equipment
`Dead Reckoning between Beacons
`Accuracy Updated at Beacons
`Satellite Positioninq
`High Vehicular Expense
`Requires Satellite Service
`High Accuracy when Reception Good
`Dead Reckoning Backup Required
`Map Matchins
`Moderate Vehicular Expense
`Requires Map Data Base
`High Accuracy when Map Correct
`Compensates Dead Reckoning Error
`
`354
`
`IPR2017-02022
`Unified EX1013 Page 5
`
`

`

`rechner" . The Nissan
`
`examples of
`Dead Reckoning Recent
`dead reckoning systems include the Nissan
`"Driver Guide", the Honda "Electro Gyro-
`Cater", and the
`Daimler-Benz "Routen-
`system (26) uses
`magnetic compass and odometer signals to
`continuously compute the distance and
`direction to a destination whose coordi-
`nates are input by the driver. A display
`comprised of an array of symbolic indica-
`tor lights shows the current direction to
`the destination, and a bar
`graph shows
`remaining distance.
`The Honda system (27) uses a helium
`gas-rate gyro and odometer to compute the
`vehicle's path relative to its starting
`point.
`The path is displayed on a CRT
`screen behind a transparent map overlay of
`appropriate scale. Provision is included
`for manually adjusting the map position to
`keep it in registration with vehicle path.
`The Daimler-Benz
`system has two
`modes, one for city and one for highway
`driving (28). When in the city mode, this
`system operates much like the Nissan sys-
`tem, using magnetic compass and odometer
`inputs to compute and display distance and
`direction to a driver-specified destina-
`tion.
`When in the highway
`mode, the
`system makes use of stored digital map
`data for the trunk highway network.
`The
`driver inputs both origin and destination,
`and the
`system computes the optimum
`highway route and prompts the driver step-
`by-step over the route.
`The next route
`point (entrance, exit, rest house, etc.)
`and its current distance is shown on a
`small alphanumeric display. Only odometer
`input is used in the highway mode, and the
`driver may manually correct any distance
`error when arriving at route points.
`Proximity Beacon This approach, now
`inactive in the U. S., has been the sub-
`ject of further development and testing in
`Japan and West Germany. The Japanese CACS
`(Comprehensive Automobile Traffic Control
`System) project (14) was essentially simi-
`lar to ERGS (16) with two-way communica-
`tions as equipped vehicles pass roadside
`units.
`CACS, which included
`linkage
`between the roadside units and a traffic
`control center for dynamic routing capa-
`bility, was tested extensively through
`large scale road and simulation experi-
`ments. Although the CACS project ended in
`1979, related studies have continued (291,
`and roadside units still have a potential
`role in advanced vehicle navigation
`systems now being planned in Japan (30).
`The West German
`ALI Destination
`Guidance System, designed primarily
`for
`freeway guidance, was developed
`by the
`Blaupunkt/Bosch group in the late 1970's
`(14). ALI uses induction loops at road
`intersections as a means of collecting
`
`data (type of car, speed destination,
`etc.)
`from equipped vehicles.
`The
`collected destination data are fed into a
`central
`computer
`where
`individually
`recommended routes are optimized with
`respect to current traffic situations.
`Route guidance information is then
`transmitted to individual vehicles and
`presented in visual form such as "turn
`left" or 'turn right". Although proven to
`work very well in a 1979-1982 field test
`covering about 150 km of the Autobahn
`network, ALI was not considered for
`general introduction because of the high
`cost of installing appropriate roadway
`infrastructure (31).
`Map Matchinq The first map matching
`system, an automatic route control system
`for delivery vehicles ( 5 1 , was limited to
`planned routes programmed before an
`equipped vehicle was dispatched. Route
`data tapes were prepared from central
`computer files, or by driving a route with
`an equipped vehicle to "record" the
`An on-board computer analyzed
`route.
`differential odometer signals, deduced the
`vehicle's path and compared it with the
`programmed path to maintain accuracy, to
`identify locations €or issuing route
`guidance to the driver, and to prompt the
`throwing of newspapers to subscriber
`houses included on the route tapes. Route
`guidance was initially accomplished
`by
`activating pre-recorded audio messages. A
`subsequent version used a plasma panel to
`display graphic instructions and explan-
`atory text (32). This system was in daily
`revenue service during a one year test by
`the F o r t Worth S t a r Telegram ( 3 3 ) .
`The first commercially available
`automobile navigation system based on map
`matching technology is the
`Etak "Navi-
`gator-'' marketed in the San Francisco and
`Los Angeles areas (the first areas with
`available map data) since 1985. The Etak
`system uses a flux-gate magnetic compass
`as well as differential odometry €or dead
`reckoning input, and uses tape cassettes
`The
`to store digital map data (11).
`system continuously displays the vehicle's
`location relative to its surroundings on a
`CRT map. A stationary arrowhead symbol in
`the center of the
`CRT represents the
`vehicle position, and points to the top of
`the display indicating vehicle heading.
`As the vehicle is driven, the map rotates
`and shifts about the fixed arrowhead
`accordingly. Input destinations are shown
`on the Etak screen as a flashing star.
`Map matching systems are also being
`developed by European developers using the
`compact disc for storage of massive map
`A pioneering
`AVL
`data bases ( 3 4 , 35).
`system for the St. Louis police fleet com-
`bined dead reckoning on the vehicles with
`map matching at the dispatch office (36).
`
`355
`
`IPR2017-02022
`Unified EX1013 Page 6
`
`

`

`The Navstar
`Satellite Positioning
`CLASS
`GPS system was the basis for
`Satellite
`(Chrysler Laser Atlas and
`System), a concept displayed at the 1984
`World's Fair in New Orleans ( 20). CLASS
`included a nationwide set of maps stored
`on video disc, and software for automat-
`ically selecting and displaying on a color
`CRT the map area incorporating the
`vehicle's current location as indicated by
`a cursor. Touch screen controls permitted
`the driver to input a destination or zoom
`in on a
`particular map area. Actual
`implementation of CLASS requires dead
`reckoning for navigating when signal
`reception is impaired by buildings,
`terrain features, etc.
`General Motors has experimented with
`GPS in an automobile system which includes
`Loran and dead reckoning (19). Ford
`has
`also considered GPS for automobile naviga-
`tion (18), and Rockwell International has
`outfitted a Collins shuttle bus with a GPS
`receiver coupled with
`an electronic map
`display (37). The bus location is indi-
`cated by a small white square on the CRT
`screen; when the vehicle is
`moving, the
`square moves across the map and leaves a
`blue trail
`to indicate the
`route trav-
`elled.
`Traffic Manaqement Automobile navi-
`gation systems with routing or guidance
`functions are potentially more useful if
`current information on traffic conditions
`is available for consideration in routing.
`Current information
`may be communicated
`with suitable data links so that routing
`and guidance generated by on-board systems
`can avoid congested areas, thus contribut-
`ing to balanced traffic management. Such
`communication links may also provide up-
`dates (new construction, detours, etc.) to
`digital map data bases used with automo-
`bile navigation systems.
`
`available,
`While the technology is
`the necessary coordination between govern-
`ment and industry to develop communica-
`tions standards and to implement and
`operate interactive traffic
`management
`systems has not been established in the
`U. S. (38). Both West Germany
`(31) and
`Japan (30), however, have formed working
`groups and committees involving road
`agencies, automobile manufacturers, elec-
`tronic device manufacturers, etc., to
`coordinate plans and standards for com-
`municating real-time traffic
`data and
`digital map updates to on-board systems.
`Interactive traffic
`management was
`recently studied by the COST 30 bis Work-
`ing Group 1 (Road/Vehicle Electronic Com-
`munication) of the OECD. The objective of
`the international group was to determine
`what aspects of a road-to-vehicle/vehicle-
`
`to-road electronic communication link
`should be standardized among European
`countries for traffic control and route
`guidance. The final report noted that the
`use of such communication links "could
`produce quantifiable benefits for drivers
`and for the community worth around €4,000
`m per year in Europe, " and recommended
`using area broadcasting
`techniques to
`transmit updates for "vehicle-borne navi-
`gation aids which should soon become
`available on the market" (39).
`CONCLUSIONS
`There are several viable technologies
`and systems approaches to high-technology
`automobile navigation, each having its own
`strengths and weaknesses. Although
`it is
`premature to conclude which combinations
`will prevail, trends are evident and major
`issues may be identified.
`One important trend
`is toward "map
`intelligent" systems that navigate
`by
`associating observed mathematical features
`of the vehicle path with those encoded in
`a map data base, analogous to the con-
`ventional driver who navigates by associ-
`ating observed landmarks and road features