Eighth International Conference on
`
`AUTOMOTIVE
`
`E LECT R0 N I cs
`
`23 - 3| October |99I
`
`Organised by the
`
`The Computing and Control and Electronics Divisions of the Institution of
`Electrical Engineers and the Automobile Division of the Institution of
`Mechanical Engineers
`
`in association with the
`
`j"
`
`'
`
`‘. «
`
`Associazione Elettrotecnica ed Elettronica ltalia
`
`Institute of Electrical and Electronics Engineers Inc
`(Vehicular Tech nology Society)
`Institute of Measurement and Control
`
`Institute of Physics
`Road Transport Forum
`Société des Electriciens, des Electroniciens et des Radioelectriciens
`
`Society of Automotive Engineers Inc
`Society of Automotive Engineers of Japan
`Society Of Motor Manufacturers and Traders Ltd
`Svenska Elektroingenjorers Riksforening
`Svenska Mekanisters Riksforening
`Verband Deutscher Elecktrotechniker
`
`Verein Deutscher Ingenieur
`
`Venue
`
`The Institution of Electrical Engineers, Savoy Place, London.WC2. UK
`
`TOYOTA EX. 1008, p. 1
`
`TOYOTA Ex. 1008, p. 1
`
`

`

`Author Disclaimer
`
`"While the author and the publisher believe that the information and guidance given in this work are correct. all parties must
`rely upon their own skill and judgment when making use of it. Neither the author nor the publisher assume any liability to
`anyone for any loss or damage caused by any error or omission in the work. whether such error or omission is the result
`of negligence or any other cause. Any and all such liability is disclaimed."
`
`Copyright and Copying
`
`“All Rights Reserved. No part of this publication may be produced. stored in a retrieval system. or transmitted in any form
`or by any means - electronic. mechanical, photocopying, recording or otherwise - without the prior written permission of
`the publisher."
`
`Conditions of Acceptance of Advertisements
`
`The institution reserves the right to refuse to insert any advertisements (even if ordered and paid for) and/or to make
`alterations necessary to maintain its standards.
`
`It is not guaranteed that any advertisement will be placed in any specified position or on any specified page unless agreement
`has been entered into and the agreed surcharge paid.
`
`Every effort will be made to avoid errors but no responsibility will be accepted for any mistakes that may arise in the course
`of publication of any advertisement. These mistakes may include non-insertion. insertions other than those ordered and
`errors and omissions within the advertisement.
`
`Notice to cancel any advertisement must be received in writing ten days prior to its next scheduled appearance. otherwise
`a charge will be made.
`
`No responsibility will be accepted for repetative errors unless the advertiser's correction has been accepted in respect of
`that error.
`
`No responsibility will be accepted for loss or damage alleged to arise from errors within advertisement copy, non-appearance
`of an advertisement or delay in forwarding box number replies.
`
`Advertisers are required to ensure that the content of advertisements conforms with all legislation currently in force
`affecting such matters. They shall further indemnify the publisher in respect of any claims. costs and expenses that may
`arise from anything contained within the advertisement and published on their behalf by the Institution of Electrical Engineers.
`
`The placing ofan order or contract for insertion ofan advertisement in any newspaper or jou rnal published by the Institution
`of Electrical Engineers whether in writing or by verbal or telephone instructions will be deemed an acceptance of each and
`all the above conditions.
`
`!
`'
`
`.
`1
`'
`
`I
`
`i
`
`.
`
`Published by the Institution of Electrical Engineers. London. lSBN 0 85296 525 7 ISSN 0537-9989.
`
`This publication is copyright under the Berne Convention and the International Copyright Conveno'on.
`All right: reserved. Apart from any copying under the U.K. Copyright. Designs and Patents Act I988. Part I. Section 38. whereby a single copy of an artide
`may be supplied. under certain conditions. for the purposes of research or private study. by a library of a class prescribed by The Copyright (Librarians and
`Archivists) (Copying of Copyright Material) Regulations I989: Sl l989/l2 I2. no part of this publiation may be reproduced. stored in a retrieval system or
`transmitted in any form or by any means without the prior permission of the copyright owners. Permission is. however. not required to copy abstract: of
`papers or articles on condition that a full reference to the source is shown.
`
`Multiple copying of the contents of the publication without permission is always illegal.
`
`© l99l The institution of Electrical Engineers
`Printed in Great Britain by Short Run Press Ltd. Exeter.
`
`TOYOTA EX. 1008, p. 2
`—
`
`TOYOTA Ex. 1008, p. 2
`
`

`

`Organising Committee
`
`Mr] P Cousins. AB Automotive Electronics Ltd (Chairman)
`Dr P A Bennett. Centre for Software Engineering
`Mr P Day. Plessey Research
`Dr B Edwards. Lucas Automotive Ltd
`
`Dr W] Gillan, Department of Transport
`Mr K W Huddart. Traffic Engineering Consultant
`Mr C Jones. Autocontrols Ltd
`Mr j Moore. Lucas Automotive Sensors
`Mr A C N Wilkinson. Ford Motor Company Ltd
`Mr M Williams. jaguar Cars
`Mr] Wood, MIRA
`
`Corresponding Members
`
`Mr E Ferrati. Italy
`Mr F Heintz. Germany
`Mr W R Kissel, USA
`
`Professor E Panizza. Italy
`Mr R R Smisek, USA
`
`I
`
`TOYOTA EX. 1008, p. 3
`
`TOYOTA Ex. 1008, p. 3
`
`

`

`Contents
`
`The Institution of Electrical Engineers is not, as a body, responsible for the
`
`opinions expressed by individual authors or speakers
`
`Page No.
`
`KEYNOTE SESSION
`
`|
`
`7
`
`'Car electronics - key factors of success for the'90s'
`W Ziebart
`
`Bayerische Motorenwerke AG. Germany
`
`'System integration: a North American perspective'
`C J Longtin
`HT Teves America, USA
`
`|
`
`|
`
`'Automotive Electronics - a Japanese perspective'
`N Miura
`
`Nissan Motor Co Ltd. Japan
`
`l9
`
`24
`
`POWERTRAIN MANAGEMENT -
`
`COMPONENTS DEVELOPMENT
`
`'Development of tin dioxide based exhaust sensors'
`P G Eastwood, T C Claypole and J Watson
`University College Swansea, UK
`
`'A capacitance sensor for methanol ratio measurement of blended gasoline'
`K Takeuchi and T Kita
`
`Nissan Motor Co Ltd. Japan
`H Kamioka. M Shimamura and K Kobayashi
`japan Electronic Control Systems Co Ltd. Japan
`
`29
`
`'High performance mixed analogue and digital ASIC for knock detection'
`K Cockerham
`
`Lucas Engine Management Systems. UK
`
`POWERTRAIN MANAGEMENT -
`
`SYSTEMS DEVELOPMENT
`
`34
`
`'A novel distributor-less variable spark energy IC engine ignition system'
`P A Howson, N T Vu and R Miller
`
`Brighton Polytechnic, UK
`F Lauerrinl and M RaJkovic
`Champion Spark Plug SA, Belgium
`
`TOYOTA EX. 1008, p. 4
`
`
`TOYOTA Ex. 1008, p. 4
`
`

`

`Contents
`
`Page No.
`
`39
`
`43
`
`'A totally integrated electronic control system for the engine and transmission'
`Y Ohyama. M Ohsuga. T Nogi and T Minowa
`Hitachi Ltd. japan
`
`'Electronic boost pressure and knock control system for S.l. engines with turbo
`charger'
`H-M Miiller, S Unland and W Himing
`Robert Bosch GmbH, Germany
`
`48
`
`'Advanced control of gasoline injection system'
`T lnui
`
`Hitachi Ltd, Japan
`
`ADVANCES IN PROCESSORS AND ASICs
`
`53
`
`'A floating point co-processor for real-time fault detection and isolation
`in electronically controlled IC engines'
`T Yu
`
`The University of Michigan. USA
`G Rizzoni
`
`The Ohio State University. USA
`
`58
`
`63
`
`68
`
`74
`
`79
`
`'Microcoded timing coprocessor for enhanced engine management'
`R Soja
`Motorola Semiconductors Ltd. UK
`
`An enabling technology for automotive systems - LinASlC'
`R Kerslake and S Oxley
`Texas Instruments. UK
`
`EMCIMULTIPLEXING
`
`'Testing of automotive electronic components regarding influence of electromagnetic
`field strength levels radiated by lightning discharges - application to a multiplexing
`communication system'
`S Ficheux
`UTAC, France
`
`M Klingler and M Heddebaut
`INRETS. France
`
`'Class 2: General Motors' version of SAE jl850'
`C A Lupini, T] Haggerty and T A Braun
`Delco Electronics Corporation. USA
`
`'Cost-efi'ective multiplexing'
`] Chidlow and P Hollins
`Lucas Rists Wiring Systems. UK
`
`vi
`
`TOYOTA EX. 1008, p. 5
`
`TOYOTA Ex. 1008, p. 5
`
`

`

`Contents
`
`Page
`
`MATHEMATICAL MODELLING AND SIMULATION
`
`85
`
`'Use of simulation in the design of automotive electronics'
`W A Havranek and A Goucem
`
`Rapid Data Ltd. UK
`
`90
`
`'A non-linear mathematical engine model for the development of dynamic engine
`control'
`M Md Ramli and A S Morris
`
`University of Sheffield. UK
`
`SOFTWARE, QUALITY AND RELIABILITY
`
`95
`
`'An overview of in-vehicle control systems development'
`M A Burchett
`
`Research and Engineering Centre, Ford Motor Company, UK
`
`98
`
`'lssues in the validation and verification of vehicle software'
`
`B Overton, | Spalding and M Thomas
`Praxis Warwick Ltd, UK
`
`|02
`
`'Managing the risks associated with software in vehicle programmes'
`T Moon
`
`AB Automotive Electronics Ltd, UK
`
`|06
`
`‘Software reliability - the link with EMC'
`K L Longmore
`Lotus Engineering. UK
`
`PROMETHEUS DRIVE
`
`1 l2
`
`'Contribution to a future European traffic structure'
`H-P Glathe
`Prometheus Office
`
`c/o Daimler-Benz AG, Germany
`
`I20
`
`I25
`
`'Digital beacon vehicle communications at 6| GHz for interactive dynamic traffic
`management'
`H-l Fischer
`Telefunken Systemtechnik GmbH, Germany
`
`'Characterisation ofa data transponding link for road-use debiting systems'
`S E ljaha
`University of Newcastle upon Tyne. UK
`E Korolkiewicz
`
`Polytechnic of Newcastle upon Tyne, UK
`
`vii
`
`TOYOTA EX. 1008, p. 6
`
`
`TOYOTA Ex. 1008, p. 6
`
`

`

`
`
`Contents
`
`Page No.
`
`I30
`
`'The use of transponder technology in road traffic control'
`P T Blythe
`University of Newcastle upon Tyne. UK
`
`I35
`
`'DRACO. A transient recorder for road accidents'
`W Fincham
`
`Queen Mary and Wesdield College, UK
`M Fowkes
`MIRA, UK
`P Ghibaudi
`
`LABEN, Italy
`
`I40
`
`'Equipment and methods for the evaluation of driver performance in relation to
`vehicle instrumentation'
`
`A Stevens and] F Collins
`Transport and Road Research Laboratory, UK
`
`I45
`
`'8OGHz automotive radar'
`A G Stove
`
`Philips Research Laboratories. UK
`
`I50 Millimetre wave technology for collision avoidance and cruise control'
`P L Lowbridge
`GEC Plessey Semiconductors, UK
`P M Brigginshaw and B Kumar
`GEC Marconi Ltd. Hirst Research Centre. UK
`
`I55
`
`I60
`
`I65
`
`'A semi-realistic driving simulator based on a video disc'
`D A Fraser, A Davis. R E Hawken, A Tollyfield, P Neave and V Seivey
`King's College, University of London, UK
`
`'Safe RTI systems - a proposal for a standard'
`P H jesty, T F Buckley and M M West
`University of Leeds, UK
`
`'The application of artificial neural networks to provide safe access to driver
`information systems and other non-critical automotive functions'
`M L Vaughn. H Van Schalkwyk and R A King
`Royal Military College of Science (Cranfield), UK
`
`I70
`
`'Driver information systems - who wants them?‘
`C Querée
`
`MVA Systematica, UK
`
`viii
`
`TOYOTA EX. 1008, p. 7
`
`
`TOYOTA Ex. 1008, p. 7
`
`

`

`Contents
`
`Page No.
`
`I74
`
`I78
`
`I82
`
`|87
`
`I92
`
`I98
`
`PACKAGING AND MANUFACTURINGTECHNOLOGY
`
`'Performance of new silicone adhesives and encapsulants at high and low temperatures'
`j-P Mollie and R L Paquet
`Dow Corning Europe, Belgium
`
`'Automated thermographic inspection of surface mount solder joints'
`P C M Hardess and D C Whalley
`Loughborough University of Technology. UK
`
`'Multichip packages - the next stage of integration for automotive electronics?I
`S Oxley
`Texas Instruments, UK
`P Prevot
`
`Texas Instruments, France
`
`COMPONENTS IN CHASSIS MANAGEMENT SYSTEMS
`
`'Vehicle auxiliary power applications for solar cells'
`I F Garner
`Solems SA. France
`
`'ROVAT - A most versatile and cost—effective brushless angle sensor'
`D L Hore and R A Slade
`Radiodetection Ltd. UK
`
`'Development of an automotive accelerometer for advanced chassis management
`applications'
`I] Harvey and M] Bliss
`Lucas Automotive Sensors, UK
`
`203
`
`'Non-contact rotary sensors for automotive use'
`S A Hale
`
`Lucas Automotive Sensors, UK
`
`CHASSIS MANAGEMENT SYSTEMS
`
`208
`
`'The ECU of a rear wheel steering system'
`H Bischof, B Donhauser and K Meder
`
`Robert Bosch GmbH. Germany
`
`2I4
`
`'Study on a new four-wheel-steering control method at low speeds - front-end path
`memorizing method'
`K Adachi, K Ito and T Fujishiro
`Nissan Motor Co Ltd, Japan
`
`ix
`
`TOYOTA EX. 1008, p. 8
`
`TOYOTA Ex. 1008, p. 8
`
`

`

`135
`
`DRACO.
`
`A TRANSIENT RECORDER FOR ROAD ACCIDENTS
`
`H. Finchaiti
`
`*H. Fowkes
`
`‘P. Ghibaudi
`
`Queen Phry & Hestf‘ield College UK; *M'IRA UK; ~LABEN Italy
`
`tag
`
`.1“!
`
`”utnu:-
`
`sixty seconds of pre«impact data, one second of impact data and
`twenty seconds of post-impact data. Data collection rates will
`need to be highest during the impact phase when vehicle
`acceleration and deceleration values are the highest.
`.
`.
`
`Crucial to successful operation of an accident data recorder is the
`reliable detection of the accident. Some mechanism must be
`defined to enable the recording function to be acrivated at an
`appropriate point. This will most likely rely upon the detection of
`an impact signal characteristic of an acceleration sensor.
`EX] T N
`ID NTRE
`RDER
`
`A review of sixteen existing accident recorders was made in order
`to determine whether any useful lessons could be learnt from
`previous devices. This review looked at recorders for use in road
`vehicles and aeroplanes. Road vehicle accident recorders were
`more numerous. however none of the recorders were for general
`commercial use. They were all scientific tools for research
`purposes.
`
`It is only in the last two years that prototype recorders have been
`developed for potential mass market application to road vehicles.
`What follows is a summary of the review to describe those features
`from which the DRACO concept evolved.
`
`Flight recorders
`
`These were the first type of crash recorders to be designed and
`installed and the information yielded has proven of undoubted
`'value with regard to safety issues. Economically and technically
`these devices are sophisticated systems that have to function and
`survive under particularly exacting environmental conditions. As
`an aircraft accident often occurs at relatively high speeds and with
`lots of fuel on board. the force and heat survivability have to be
`excellent The fact that an aircraft crash is a very rare occurrence
`also makes it necessary to acquire as much data as possible from
`every crash. The large number of control and communication
`systems working in an aircraft also makes it necessary to provide a
`large number of input channels and large storage capacities.
`
`Normally. for large aircraft. two separate recorders are installed:
`- a Cockpit Voice Recorder (CVR) which records all oral
`communications for the last 30 minutes.
`In addition there is an
`area microphone in the cockpit which records background noises.
`such as the click of a switch.
`- a Digital Flight Data Recorder (DFDR) which typically monitors
`data of between 5 and 100 different parameters and stores these in
`a time frame ofduration 25 hours.
`
`A tape loop is used as the storage medium. This is plastic—based
`and has been specially treated to help withstand temperatures up to
`100 degrees centigrade.
`'Ilie recorders are encased in titanium.
`steel or an alloy of the two which must withstand a temperature of
`1100 degree centigrade for 20 minutes. There are two
`encapsulations with an insulating layer between them.
`i
`n r
`r f r r
`v hi
`I
`
`The earliest such recorder we have traced was dated 1972 (EG&G.
`USA). It was also the cheapest at $12. This was only available as
`a prototype and had the prime function chime-resolving
`acceleration data on two axes. The recording device consisted of a
`spring/mass pendulum. an LED light source pointing at a concave
`minor which was a pan ofrhe inertial mass. Photographic film
`was used as the recording medium. The lower end of the
`pendulum was fastened rigidly to the vehicle body. After
`developing the photographic film. the excitation and direction of
`
`W T
`
`he major economic and social benefits of road transport have to
`be weighed against some significant penalties. The most telling of
`these is the toll of approximately one and a half million people
`killed and injured throughout the EEC each year. Any actions
`taken to mitigate this major problem rely upon an understanding of
`how and why accidents happen. This understanding is developed
`from both manual collection of accident information and in—depth
`accident research. Clearly if some more automated. global means
`of collecting accident information were available then it could aid
`our undersranding of the accident process. The EEC has identified
`this area as one wonhy of consideration and hu comrnisioned a
`consortium to evaluate the concept under the DRIVE initiative.
`This paper describes a transient recorder for road vehicles. It is
`named DRACO (DRiving Accident Coordinating Observer).
`
`An accident recorder has the basic function of collecting data
`which relates to the behaviour of a vehicle when it is in an
`accident. In this sense it simply acts as an on»vehicle observer of
`that vehicle's movements and internal systems only when it has
`been subjected to the transient event known as an impact. The
`recorded data can then be used in the process of accident
`reconstruction.
`
`Accident reconstruction is a formal procedure to determine the
`events which lead to an accident by analysis of data collected from
`the accident site. Such information might be used in a criminal or
`civil action in a court of law.
`
`NERAL RE
`
`IREMENT
`
`F AN A
`
`I
`
`RDER
`
`W I
`
`n the case of road vehicle accidents. the most important data
`required will be the closing specd(s) and positions of the vehicle(s)
`immediately prior to impact. Currently. the usual procedure to
`determine this information is to use measurements and other
`information obtained by accident investigators called to the scene
`after the event. Observation of skid marks and vehicle damage
`now provide the basis for the process known as accident
`reconstruction. Computer based algorithms. or equivalent manual
`methods. may be used to assist in the estimation of the most likely
`vehicle movement trajectory (position-velocity profile) and these
`are based on the momentum exchange and energy absorption
`principles of classical dynamics.
`
`An accident data recorder will need to record parameters that
`would enable the physical dynamics of the vehicle to be
`reconstructed in post-accident analysis. This should include
`acceleration. velocity and vehicle attitude. In order to facilitate the
`reconstruction process any acceleration data would need to be
`integrated once to provide velocity data and then again to find
`posuion. Naturally. reliable initial conditions are required to
`Complete the reconstruction. These will generally be. in fact. final
`COnditions relating to where the vehicle comes to rest i.c. the zero
`velocity state. The implication of this procedure is that an accident
`data recorder such as DRACO must also record time accurately.
`
`It would also be useful to record some status signals from common
`Vehicle systems. such as lighting and braking circuits. to be aware
`of their functional state during the accident. and to enhance the
`understanding of the event.
`' n
`r
`r in
`
`Examination of accident events and other recorders suggest that
`adoquate information for reconstruction will be provided from a
`dam wrndow of approximately two minutes subdivided as follows:
`
`TOYOTA EX. 1008, p. 9
`
`
`TOYOTA Ex. 1008, p. 9
`
`

`

`136
`
`’—.—._...—-——__...
`
`__.._..__.
`
`can also cause a failure of the vehicle power supply. It is perhaps
`necessary to provide for a memory back-up battery within the
`recorder for data retention purposes.
`It is essential to provide data
`recording until the vehicle has come to rest after the crash.
`,
`.
`
`This is a necessity for a system which is supposed to be used for
`accident reconstruction and perhaps for the settlement of liabilities.
`As soon as a vehicle speed signal is connected to a recorder there
`is the likely temptation amongst some vehicle users to somehow
`tamper with the recorder to try to ensure that a lower value than
`the actual speed is recorded. Manufacturers of tachographs have
`developed methods whereby any tampering can usually be
`detected.
`
`5'!"
`
`Some Heel Management System (FMS) data recorders have a self-
`test routine which provides a warning to the driver of malfunction.
`Such a feature is important to an accident recorder especially if its
`use is to be mandatory.
`
`These considerations formed part of the influences on the
`development on the DRACO concept. This is described below
`
`TllE QBACQ UH“
`
`DRACO is conceived as an on-board vehicle accident data
`recorder. The recording function is only activated when an
`'accident' has been detected by an accident detection algorithm
`(ADA) present in the unit. This ADA function is constantly
`monitoring several data channels for this purpose.
`
`DRACO will be capable of recording data from both inbuilt
`sensors within the unit and from some sources from within the host
`vehicles electronic/electrical systems. The data channels selected
`are X. Y and Z axis accelerometers. yaw (ie rotation about the Z
`axis). vehicle speed. status signals concerning vehicle external
`lights (side. head. brake and indicator lights). horn and also the
`time of the detected accident.
`
`DRACO will be a solid-state micro-processor based unit fitted
`with three accelerometers and a yaw rate sensor. it will require a
`standardised electrical/electronic interface to the vehicle system
`
`DRACO will be housed within a package of maximum size
`100mm x 60mm x 60mm and maximum weight 500 grams. It will
`be mechanically mounted to the vehicle structure. within the
`passenger compartment. to a defined rigid mounting surface.
`Removal of the unit from this mounting will be by specialised
`extraction tool to minimise tampering.
`
`DRACO will have capacity for 5 accident events in its memory
`and this memory and the clock will be maintained by internal
`battery back~up for up to 4 weeks after vehicle supply has failed.
`
`DRACO data can be downloaded only via a sealed serial port
`separate from the data input interface. again to minimise
`tampering.
`
`DRACO will have to continue to function under normal
`automotive durability and environmental conditions. The principal
`area of application will be the pusenger car based vehicles
`(category M1 and N1) with alternative configurations of DRACO
`for other categories of road vehicles.
`
`W T
`
`he DRACO project has. as part of its output. the construction of a
`Laboratory Model (LM). his is not intended to be a full
`prototype but will be a working recording system to be used for
`important aspects of performance assessment within a laboratory
`environment i.e. it is not intended to be fitted into a vehicle. The
`main purpose of the LM will be to develop at least one system
`architecrure which will enable a stated performance objective to be
`attained. The LM is thus an evaluation and validation tool using
`solid-state technology. with the possible exception of any sensorS.
`which will supply many relevant inputs to the later design activm'
`for DRACO highlighting the most critical system areas and glVlflg
`possible solutions to optimise the system reliability. The LM
`
`force could be read. Since the film was not mounted spherically.
`the curve had to be corrected through a lens system.
`
`In 1977 the Breed Corporation. USA. produced a recorder which
`registered the total velocity change related to die crash. The basic
`functional principle of this relatively mechanical simple device is
`that the movement of a body connected to a viscous damper can be
`made to approximate. usefully, to an integrator.
`
`More recently, Mannesmann Kienzlc. in close cooperation with the
`VOLVO Car Corporation. developed and tested a crash recorder to
`be installed in the VOLVO vehicle test fleet. The objective of the
`project was to gain experience concerning the correlation between
`crash severity. vehicle damage and bodily injuries of the occupants
`by getting access to extensive in-vehicle data from real-life
`accidents.
`In addition to acceleration values and yaw. this device
`records the status of the brake light. direction indicators. ignition
`and also measures seat belt force. The amount of data stored
`corresponds to an approximate time frame from —10 s to +10 5 with
`0 sec being the impact time. The automatic accident detection
`function is based on a programmable trigger-algorithm which
`averages deceleration values over time.
`
`Mannesmann Kienzlc subsequently developed an accident data
`recorder for road vehicles, designated the ADR-2165.
`It became
`available for pilot use in the first half of 1991. This is mainly
`destined for use in passenger cars and can be installed by the
`vehicle owner himself. This device uses accelerometers and a yaw
`sensor within the recorder and the data recorded in an accident can
`be evaluated by computer-aided accident reconstruction.
`
`Contemporary with the Mannesmann Kienzlc recorder is the
`BMFI‘IARGE Crash Recorder whose development was sponsored
`by the Bundesministerium fur Forschung und Technologie
`(BMF‘I‘) working with Kolley und Partner in Gemtany. Again the
`objective was to develop a crash recorder for private cars.
`lnfon-nation regarding this project is. however. very sparse.
`
`The only mandatory type of recorder is the tachogtaph which is a
`compulsary fitment for heavy goods vehicles (HGV) within the
`EEC.
`In this. the vehicle speed is graphically presented on a
`circular chart as a function of driving time. Although not intended
`as an accident recorder this can be used to obtain directly the speed
`of the vehicle before an accident. Although the disc diameter is
`only 12 cm for a 24 hour recording period. optical enhancement
`can produce a time resolution of 1 sec.
`
`angrgl observations
`
`Im a
`
`i
`
`n
`
`i n
`
`Detection ofan accident event will be facilitated by detection of a
`characteristic change in vehicle acceleration as registered by the
`recorder. A parallel here would be the trigger devices associated
`with air-bags or seat-belt pretensioncr deployment systems.
`In
`these. the device must operate within a few milliseconds of the
`high-g phase of an impact being initiated. This is achieved via a
`trigger algorithm which examines the deceleration values over a
`short time window to see if a certain threshold value is exceeded.
`
`i nal con iti nin
`
`an
`
`ata m
`
`i n
`
`An in-vehicle accelerometer signal inherently contains significant
`high frequency noise from vibrations of the vehicle body. These
`signals should be suitably filtered before any processing is carried
`out to provide a high noise immunity for the accident detection
`function. From the crash recorders researched one notes that when
`monitoring the impact using accelerometers. a sample rate of
`approximately 500 Hz is necessary. For the pre-crash phase a
`sample rate, of one-tenth this rate should be acceptable for accurate
`reconstruction.
`
`P'k'in'
`
`wr
`
`This must be designed to protect. at least. the data storage
`components against damage by high-g forces. din. liquids and
`excessive heat.
`It should also help to provide a certain degree of
`EMC-ptotection for the internal electronics. An accident data
`recorder must be securely fastened in a mechanically well-
`protected place such that the measurements of the internal
`accelerometers will be meaningful. For convenience. the wiring
`harness of the vehicle should be within easy reach. An accident
`
`TOYOTA EX. 1008, p. 10
`
`
`TOYOTA Ex. 1008, p. 10
`
`

`

`'‘-"|.\1-“..JO'I
`
`‘.
`
`- accident detection through the optimal identified ADA
`algorithm;
`- compression of data from sensors (if required).
`
`A tradeoff analysis of available CPUs has lead to selection of the
`Siemens SAB 80C535 8-bitconu’ollcr. lts availablity in CMOS
`allows low power consumption.
`
`W T
`
`o suppon the accident information recording capability a large
`data—recording section is needed which is to be also of non-volatile
`type. A tradeoff analysis identified battery backed static RAM and
`Flash EEPROM as being suitable devices. Both these memories
`were included into the Laboratory Model system to evaluate their
`performances. Tests have shown that the second of these devices
`is the most useful for DRACO. Indications are that 1 to 4 Mbit of
`storage is required.
`
`The main features of a Flash device is its ability to preserve the
`stored data without any power supply and the possibility of writing
`and erasing its bytes. This last feature is managed through the
`application of 12V to the chip: without it the device is a read-only
`memory. Unfortunately the memory can be only block-erased and
`hence to preserve part of its contents from an erase operation it is
`necessary to use a second back-up device.
`
`The use of a second CPU to drive the storage section is essential
`for its data file management together with implementation of a
`complex ADA. The memory controller must be hardware and
`software compatible with uPl in addition to being of low cost.
`CMOS and available in a PLCC package. Further it was required
`to have at least four I/O ports. at least 8KByte of internal maskable
`ROM. some tens of bytes of internal RAM and l or 2
`timer/counters. An lntel 80C518H belonging to the lntel MCS-Sl
`family and the Siemens SAB 8052 belonging to the Siemens
`805 l x family were selected as possible alternatives.
`
`The interfagg mtwggn P§ @351 RE
`
`The PS and RS sections are interfaced at two levels to connect
`their own processors directly through [/0 ports and through a
`volatile RAM memory shared between the two controllers. This
`port-to-port connection allows the controllers to communicate with
`each other. using a dedicated protocol. The controllers may not
`access the shared RAM at the same time and their memory R/W
`operations are software managed through their communication
`protocol.
`In any case. to prevent buffer contention problems each
`Address/Data bus to the RAM memory is decoupled. Because the
`tnapping of the RAM memory is managed circularly this allows
`uPl to down-load input data at a fixed rate and uP2 to manage
`them at different rates as required to accurately describe the
`dynamics of the accident. see Table 1 above.
`ration l
`f
`n
`rf rm
`x
`
`The LM has so far provided a good software development
`environment for the Accident Detection Algorithm which is
`reasonably representative of the expecred real situation.
`
`The mounting technology and devices selection have shown that
`the small size required for the final system is a reasonable goal.
`The whole system is cased into a metal-box of size 45x70x125mrn
`and can be connected to the outer world through a 9-pin standard
`Cannon connector which includes the power supply lines and the
`serial RS 232 line to interface the LM to a Personal Computer.
`
`Communication software (MONTERM) allows down-loading of
`assembler A5] or basic MCS 52 programs into the system; a
`Monitor/Debugger program (Monitor 51), EPROM resident.
`allows for the testing of the code developed. The most imponant
`tests implemented on the LM concerned the Flash EEPROMs. the
`A/D convener and the ADA algorithm. Results showed that:—
`
`The minimum byte programming time for the FEEPROMS.
`excluding system overhead. was l6tts (lOps program + 6145
`program verify) while the maximum was 400m (25 program
`verify loops are allowed hence 16u5‘25=400us). Typical byte
`programming time was about 40m and hence about 6s are needed
`to program 128 1(8er (one memory). The minimum chip erase
`
`137
`
`therefore embodies the outline of the DRACO unit as described
`above.
`
`We.
`
`This section outlines a Reference Architecture for DRACO to be
`used as a basis for funher detailed analysis. The architecture is to
`be primarily intended as a functional breakdown definition and so
`the building blocks depicted in Figure 1 are not necessarily well
`defined circuit sections. The three main sections are : an
`Acquisition Section (AS). :1 Processing Section (PS) and a
`Recording Section (R5).
`The A5 contains all external and internal sensors. their interfaces
`and signal conditioning circuits, the AID converter and the serial
`line. The power supply is also considered part of this section since
`it has an interface to the vehicle battery. The PS contains a Single
`chip microcomputer (uPl) as well as the required support devices
`e.g. program ROM. RAM. timers etc. The RS will be used to store
`the required accident related data. in a non-volatile form. which
`will be controlled by another microprocessor (uPZ).
`il
`f
`w
`fw
`
`The DRACO system is an intelligent acquisition and recording
`system. Figure 2 shows a block diagram of the proposed physrcal
`architecture of the unit based on two microcontrollets.
`
`At power-on. and after having run the self-test routine. uPl runs
`the Accident Detection Algorithm (ADA) which samples input
`data at a lKHz fixed rate and. after processing data. stores them
`into a volatile buffer. This is wide enough to store up to 1 sec. data
`and it is managed circularly.
`
`Synchronously with uPl. pPZ runs its own program which
`transfers data from the buffer to a volatile RAM memory at
`different rates depending on the phase (as interpreted by t