SAE TECHNICAL
`PAPER SERIES
`
`980007
`
`Adaptive Light Control - A New Light Concept
`Controlled by Vehicle Dynamics and Navigation
`
`J.P. Lowenau, J.H. Bernasch, H.G. Rieker,
`P.J.Th. Venhovens, J.P. Huber, and W. Huhn
`BMW AG Research and Development
`
`F.M. Reich
`Technical University Vienna
`
`Reprinted From: Automotive Lighting Technology
`(SP-1323)
`
`e .A I/!!! The Engineering Society
`~ For Advancing Mobility
`~ Land Sea Air and Space®
`INTERNATIONAL
`
`International Congress and Exposition
`Detroit, Michigan
`February 23-26, 1998
`
`400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A.
`
`Tel: (724) n&-4841 Fax: (724) ns-5760
`
`Mercedes EX1018
`U.S. Patent No. 11,208,029
`
`

`

`The appearance of the ISSN code atthe bottom of this page indicates SAE's consent that
`copies of the paper maybe made for personal orintemal use of specific clients. This consent
`is given on the condition however, thatthe copier pay a $7. 00 per article copy fee through
`the Copyright Clearance Center, Inc. Operations Center, 222 Rosewood Drive, Danvers,
`MA 01923 for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright
`Law. This consent does not extend to other kinds of copying such as copying for general
`distribution, for advertising or promotional purposes, for creating new collective works, or
`for resale.
`
`SAE routinely stocks printed papers for a period of three years following date
`of publication. Direct your orders to SAE Customer Sales and Satisfaction Department.
`
`Quantity reprint rates can be obtained from the Customer Sales and Satisfaction Depart(cid:173)
`ment.
`
`To request permission to reprint a technical paper or permission to use copyrighted SAE
`publications in other works, contactthe SAE Publications Group.
`
`GLOBAL MOBILITY DATABASE
`All SAE papers, standards, and selected
`books are abstracted and indexed in the SAE
`Global Mobility Database.
`
`No part of this publication may by reproduced in any form, in an electronic retrieval system
`or otherwise, without the prior written permission of the publisher.
`
`ISSN 0148-7191
`Copyright 1998 Society of Automotive Engineers, Inc.
`
`Positionsandopinionsadvancedinthispaperarethoseoftheauthor(s)andnotnecessarily
`those of SAE. The author is solely responsible for the content of the paper. A process is
`available by which discussions will be printed with the paper if it is published in SAE
`Transactions. For permission to publish this paper in full or in part, contact the SAE
`Publications Group.
`
`Persons wishing to submit papers to be considered forpresentation or publication through
`SAE should send the manuscript or a 300 word abstract of a proposed manuscript to:
`Secretary, Engineering Meetings Board, SAE.
`
`Printed in USA
`
`9 0-120 38 / PG
`
`

`

`Adaptive Light Control - A New Light Concept
`Controlled by Vehicle Dynamics and
`Navigation
`
`980007
`
`J.P. Lowenau, J.H. Bernasch, H.G. Rieker, P.J.Th. Venhovens, J.P. Huber, W. Huhn
`BMW AG Research and Development
`
`F.M. Reich
`Technical University Vienna
`
`ABSTRACT
`
`A new light concept called Adaptive Light Control (ALC)
`is developed with the aim to improve night time traffic
`safety. ALC improves the headlamp
`illumination by
`means of continuous adaption of the headlamps accor(cid:173)
`ding to the current driving situation and current environ(cid:173)
`ment. In order to ensure rapid prototyping and early
`testing
`the step
`from offline
`to online
`(real-time)
`simulation of light distributions in the driving simulator
`has been successfully done. This realtime simulation
`enabled
`the
`interactive development of new
`light
`distributions in different driving situations and driving
`environments. The solutions are directly ported to real
`vehicels
`to allow
`further testing with natural
`road
`conditions.
`
`In this paper, results of the development of movable
`headlamps are presented. These headlamps are
`controlled by means of path prediction based on vehicle
`dynamics and route vectors of the navigation system. A
`comparison of static and dynamic light distributions is
`illustrated and an outline of the basic concept of the
`Adaptive Light Control system is given.
`
`INTRODUCTION
`
`For decades, car headlamp systems have basically
`remained unchanged. The driver can select high or
`dipped headlights and perhaps switch on fog-lamps if
`necessary. These systems, of course, have become
`many times more powerful with the light bulbs of
`yesteryear replaced by halogen lamps and, ultimately, by
`xenon headlights, which are three times as powerful. Due
`to specially cut
`lenses, projection headlamps, and
`
`computer-calculated reflectors, utilization of light quantity
`has also been increased by thirty to seventy percent. But
`the headlamp system as such has remained essentially
`unchanged. The possibility to direct the available light
`dynamically into the direction where it is actually needed
`has remained a wish. In recent years, however, there has
`been a considerable amount of interest in improving the
`light quality of automotive headlamps to improve night(cid:173)
`time driving safety. A few technical realizations using
`turnable reflectors have been presented in the past (e.g.
`Ref. 1,2), and subsequent investigations now prove that
`different driving
`situations
`require different
`light
`distributions to obtain an optimal illumination (see Ref. 3-
`8). In addition, viewing strategies of the driver -especially
`during curve negotiation - are an important issue for the
`design of dynamic light distributions (Ref. 9)
`
`In the following sections we will give an outline of the
`interactive environment
`for developing new
`lighting
`systems using real-time simulation of light distributions in
`the driving simulator and the first version of ALC with
`moveable headlamps in real vehicles.
`
`INTERACTIVE ENVIRONMENT FOR NEW LIGHTING
`SYSTEMS
`
`Normally, a headlamp is first designed and then its
`characteristics are tested and optimized in practice. Due
`to the difficulty and costs of developing appropriate
`automotive headlamps, a lot of work has been done in
`light simulation. Usually static
`light distributions for
`specific situations have been modelled in an offline
`simulation, with a CPU-time from 1 O seconds up to
`minutes per frame (e.g. Ref. 10). Until now, a real-time
`simulation of (dynamic) light distributions did not exist,
`
`33
`
`

`

`light distributions
`
`• Evaluation of
`distributions)
`• Changes of the distributions (e.g. low beam and low
`beam with spot)
`• Movable headlamps with curve illumination
`
`(different
`
`light
`
`and hence, a tool to evaluate light distributions adapted
`to different driving situations or influenced by different
`weather conditions has not been available.
`
`Fig. 1: The headlamp system in the driving simulator
`
`By combining high-performance computers and specially
`developed lighting software with a real-time kernel, an
`environment is realized which allows online simulation of
`light distributions and, therefore, interactive development
`of lighting systems. The simulation produces a series of
`images (30-60 frames/second) that is projected onto a
`screen and enables one to drive interactively in the
`simulator with modelled light distributions, as
`illustrated
`in Fig. 1.
`
`DRIVING SIMULATOR - The heart of the driving
`simulator is comprised of several high-performance
`supercomputers operating in conjunction with specially
`developed lighting software. The vision simulation is
`performed by 2 closely coupled graphic engines with a 2-
`pipe and a 1-pipe machine, respectively. The graphic is
`equipped with
`two multi-channel options, six raster
`managers (16 MB) and 12 CPUs. Several graphical
`databases are implemented including textures for trees,
`buildings, signs, and cars.
`
`LIGHT COURSE AND DRIVING SIMULATION
`
`LIGHT COURSE - To test different light distributions a
`specially designed light course was developed. The
`driving simulator light course contains a closed course
`with many different curves and includes a crossing. The
`radius of curvature of the bends varies between 25 and
`150 m. The width of the road is 7 m with a center line and
`beacons every 50m. Decorated with buildings, bushes,
`trees, pedestrians and
`traffic signs,
`it provides a
`photorealistic appearance.
`
`Fig. 2: Integrated light distribution in the driving simulator
`
`ONLINE MANIPULATION OF LIGHT DISTRIBUTIONS
`
`Fig. 3: lsolux diagram at road level
`
`The light distribution is simulated not only in real-time but
`is also changable online and can be checked
`immediately. Fig. 2 shows
`the shape of cut-off
`illumination of the road that is integrated with the lighting
`software in the driving simulation (Ref. 10). Fig. 3 shows
`the isolux diagram at road level for the same distribution.
`With
`these
`light distributions, as an example,
`the
`following issues could be addressed:
`
`• Cut-off line simulation
`• Dynamic headlamp levelling
`
`34
`
`CONCEPT OF ADAPTIVE LIGHT CONTROL (ALC)
`
`During night-time driving it is more demanding to keep
`the vehicle within the lane boundaries. The automotive
`lighting equipment illuminates the driver's viewing field
`and has to deliver the appropriate information for lane
`keeping under many different driving situations.
`
`

`

`Vehicle Speed I
`
`YawAAt11
`
`Steem<1 An<le
`
`Vehicle
`Sensors
`
`VohcleSpeed
`
`Follered Side Slip Volocily
`Filered Yaw Rate
`Disturbance Lateral Oistulbance Accelaration
`Estimation
`
`R•erec1 Steemg Angle
`
`Corrector
`Predicton
`
`cx.y),.
`1
`
`Fig. 7: Structure of the dynamic path prediction
`
`The disturbance
`DISTURBANCE ESTIMATION
`estimation is based on the afore-mentioned simple ve·
`hicle model. The disturbance estimation is necessary tc
`avoid an undesired curvature of the path prediction due
`to the driver's counteracting steering motion in situations
`such as superelevated road surfaces or side wind.
`
`Fig. 8: Control of the headlamps with path prediction
`
`The equations of the well known single track model are
`known from literature (see e.g. Ref. 12 and 13). Fig. 8
`shows the movable headlamps controlled by the path
`prediction, which is identified by the two curved white
`lines.
`Fig. 9 and Fig. 10 illustrate the differences between a
`conventional low distribution (light distribution mentioned
`above) and special curve specific distribution on a normal
`country road.
`
`Experiments in the driving simulator have shown that the
`vehicle dynamics based path prediction used to control
`the headlamps is limited to a certain road geometry. It is
`insufficient when encountering s-curves, crossings, and
`other types of special road construction, because the
`driving dynamic does not know anything about the road
`geometry
`further ahead. Additional knowledge
`is
`necessary to overcome these deficiencies.
`
`Fig. 5: Driving situation with Adaptive Light Control
`
`Regarding the driver's needs for visual information about
`the
`road ahead,
`present headlight
`technology
`insufficiently lights up bends in the road ahead. The
`advantage of using Adaptive Light Control is shown in
`Fig. 5, where one observes a country road lit correctly.
`Instead of illuminating some fields or trees, the available
`light is directed dynamically in the direction where it is
`actually needed, enhancing the visual information for the
`driver.
`
`BASIC STRUCTUR OF ALC
`- The newly
`developed movable headlamps are controlled by means
`of path prediction based on vehicle dynamics and route
`vectors of the navigation system (see Fig. 6). With the
`output of the path prediction model it is possible to
`control the movable headlamps appropriately.
`
`Vehicle Dynamics I
`
`I
`
`•
`
`Headbeam Control
`
`Navigation System I
`
`I
`
`Fig. 6: Structure of Adaptive Light Control - ALC
`
`DYNAMIC PATH PREDICTION - Fig. 7 shows
`the structure of the path prediction. The prediction
`consists of an estimation of the (lateral) disturbances ( or
`accelerations) acting on the vehicle body and a prediction
`of the future path of the vehicle based on the current
`vehicle dynamic states in combination with a simple
`model of the vehicle. A filtering has been added such that
`the variations in the prediction of the future path, e.g. due
`to steering wheel motions, can be controlled (Ref. 11 ).
`
`35
`
`

`

`Fig. 9: Conventional luminous intensity distribution
`
`Fig. 10: Curve specific luminous intensity distribution
`
`NAVIGATION DATABASE - In addition to the
`vehicle dynamic parameters, route vectors from the
`navigation system are used to control the movable
`headlamps. An electronic compass, two wheel sensors,
`and GPS data are the input of the navigation system
`(Fig. 11}.
`The navigation system permanently keeps track of the
`precise location of the car, and with its built-in digital road
`map (CD-ROM} it knows the road characteristics ahead.
`The system provides position accuracy of 10 to 30 m for
`country roads and cities.
`
`SIMULATED NAVIGATION DATA - To be able
`to use the navigation system in the driving simulation the
`output data are simulated in a manner reflecting the
`characteristics known by the real system. Even a
`simulation of distorted or low quality data is possible and
`was included in the overall solution. A special protocol is
`used to provide the following information:
`• current car position
`•
`route data vectors of the scheduled route, with
`- reliability of the route data vectors,
`- current road class
`- on-map or off-map status.
`
`These data are necessary to develop control strategies
`to produce light distributions for adaptive lighting
`situations depending on the items mentioned above.
`
`GPS
`
`Compass
`
`ABS Wheel Sensor
`Sensor
`
`CD Drive
`
`Sensor lnteiiace
`
`MMI
`Display
`Speaker
`Control Panel
`
`Fig. 11: Structural diagram of the navigation system
`
`36
`
`

`

`Frame Graobet Cau:i liill!liu
`CAN Bus card ~
`lliillS:::..
`COM Controeer
`• •
`
`A/Ocard
`
`I •
`
`I Video Camera I
`
`1 v-~ .. , I
`
`Video Convener
`
`Sensors
`Vehlele Velocrty
`YawRale
`Ram
`
`l..tghl Coolrol
`System
`
`Fig.14: ALC-Hardware in the vehicle
`
`With our modular program structure we are able to test,
`tune, and evaluate ALC strategies in the simulator.
`Figure 13 shows the program structure of ALC based on
`the navigation system.
`
`For the development of ALC the components described
`in Fig. 14 are used. The ALC headlamp in Fig. 14 is
`designed to fit into existing series vehicles and to fullfil
`the ALC requirements.
`
`RESULTS
`
`Using special graphic algorithms a real-time simulation of
`light distributions with 30-60 frames/second has been
`achieved. Using this real-time simulation an adaptable
`light distribution for dynamic movable headlamps is
`developed for the light system Adaptive Light Control
`(ALC). The solution of ALC with control algorithms and
`lighting strategies has been realized at first in the driving
`simulator. It has been ported with minimal effort to real
`vehicles, the first hardware prototype of ALC is now
`under investigation. First experiments demonstrated the
`subjective advantage of ALC for the driver, it seems to
`serve as active guidance device for the driver by means
`of lighting up the scenery primarily in driving direction
`(eg. the road even in curves).
`
`OUTLOOK
`
`In this paper an outline of the environment for developing
`light distributions and movable headlamps in a driving
`simulator is presented. First results of experiments as
`well in the simulator as with real vehicles are very
`promising. In the near future the focus will be on lighting
`strategies and optimal control algorithms and on the
`interaction of eye movement behaviour and dynamic light
`patterns.
`
`Fig.12: Visualization of path prediction and navi-data
`
`DATA INTEGRATION - As a final step, the path
`prediction and the navigation data have to be coupled to
`ensure consistent control of the movable headlamps. Fig.
`12 shows both, the path prediction from the driving
`dynamics (white lines left and right) and the vector of the
`simulated navigation system (white line in the middle).
`
`Vehicle Velocity
`
`Position Data &
`Driving Direction
`
`Route Data
`
`i
`I
`I
`I
`I
`I
`I
`I
`I
`L _ _
`
`Tn,nsfonnation to Vehicle Frame
`
`Route Data Interpolation
`
`ALC Strategy
`
`ALC Light Distribution
`
`Fig.13: Interface of ALC and Navigation
`
`37
`
`

`

`REFERENCES
`
`1. Wada, K., Miyazawa, K., Yaki, S., Takahashi, K.,
`Shibata, H. (1989) Steerable Forward Lighting
`System. SAE Technical Paper Series 890682, USA.
`
`2. Kobashi, S., Hayakawa, M. (1991) Beam Controllable
`Headlighting System. SAE Technical Paper Series
`910829, USA.
`
`3. Braess, H-H., Reichart, G. (1995) Prometheus: Vision
`des intelligenten Automobils auf intelligenter StraBe,
`ATZ 95/6.
`
`4. Schmidt-Clausen, H.J., (1995/97) Progress in
`Automobil Lighting. PAL Conference, Technical
`University Darmstadt.
`
`5. Huhn, W., Hege, G. (1995) High Intensity Discharge
`Headlamps(HID).
`SAE Technical Paper Series 950591, USA.
`
`6. Hogrefe, H., Neumann, R. (1995) Adaptive Light
`Pattern - A New Way to Improve Light. SAE Technical
`Paper Series 970644, USA.
`
`7. Damasky, J. (1995) Lichtechnische Entwicklung von
`Anforderungen an Kfz.-Scheinwerfer. Dissertation, TH
`Darmstadt.
`
`8. AFS - Advanced Frontlighting Systems, (1996)
`Eureka Project No. 1403. Phase 1 - Feasibility Study,
`Final Report.
`
`9. Bernasch, J.H., Bengler K., L6wenau, J.P., Scholles
`M. (1997) Comparison of Viewing Strategies during
`Curve Negotiation in Reality and in Simulation. In
`Proceedings of the Driving Simulator Conference
`DSC '97, Lyon, France
`
`10. Bernasch, J., Haenel, S. (1995) The BMW Driving
`Simulator Used for the Development of a Driver(cid:173)
`Biased Adaptive Cruise Control. In Proceedings of the
`Driving Simulator Conference DSC '95., Sophia
`Antipolis, France.
`
`11. Reich, F.M., Bemasch, J.H., Lowenau, J.P. (1996)
`Online Steering Dynamics in the BMW Driving
`Simulator. IMAGE Conference 1996, Scottsdale,
`Arizona, USA.
`
`12. Mitschke, M. (1992). Dynamik der Kraftfahrzeuge,
`Band C: Fahrverhalten, Springer Verlag, Berlin.
`
`13. Reich F. M., Venhovens P. J.Th., Bernasch J.H.,
`Lowenau Jan P. (1997) Optimized Haptical,
`Acoustical and Visual Tuning with Different Vehicle
`Dynamic Models for the BMW Driving Simulator,
`IAVSD 15th Symposium Dynamics Of Vehicles on
`roads and tracks, Budapest, Hungary
`
`38
`
`