Home

Development of a Low-Cost Automated Crash Notification System

1. lt minruntoral mMintuntobral c nningtcobal 4 time 000 walues under 255 maxruntotal bk maxruntotal Minr ntotal bk mrinruntoLal maxruntotal maxr r n otal 1002 7 2008 upper minruntotal 16 lower fmod minruntotal 16 assign hex values output 0 reference upper output reference lower outputLl2 e t 13 upper maxruntotal 16 lower fmod maxruntotal 16 assign hex values output 3 reference upper output 4 reterence lower Output 5 T 65 write value to serial port Dwrite 21 1 Ou tp t Ser lem OuepUE Jy Dzlsend prompt time index 0 if crash 1 Dwrite 244 Coordinate strlen coordi mave Dzlsend prompt Maxrunto Leal Maxruntotal Dis minruntotal minruntotal bk fia lt T10 Xcun snobotal gt T397 5610 iit UII 5 crash 1 circbuffer index input index Cime indet 1 index 50 index 0 VOLO PRIO Logd value DI TCR xw inhibit TIMERO interrupt Outport TMDROL load value E outport TMDROH load value gt gt 8 outport RLDROL load value outport RLDROH load value gt gt 8
2. MODE BAUD NO MODEM ECHO Dinit 40 TBUESIZE MODE NO MODEM ECHO input 0 j 0 while 1 runwatch nold value OUEDOrE Coots 21 begin conversion to A D OUEDOBbEICST 0 cbutfer for new coordinates LEC Dread 3e X while j lt RBUFSIZE 1 je wait for interrupt while status 0 0100 read A D value inport C54 return status register to normal and awaiting the next interrupt status 0x0000 release held value O uEtporLi soT4d quU ys place algorithm here r umnmnrngbotal r nput circb ffer index gt 5unnipnogtoralsStlbo J 297 11 crashtransmission circbuffer index input 68 index if index 50 index 0 CALS Ini alization function for the programable interrupt ERTO init uint load value DI IDDOPPCOIOBR inhibit TIMERO interrupt outport TMDROL load value outport TMDROH load value gt gt 8 outport RLDROL load value outport RLDROH load value gt gt 8 ICR lnportopoR 8090910001 this points the interrupt vector to the interrupt handler INT VEC PRTO VEC ISR readAD th
3. 2 Figure 3 1 System 5 Figure 4 1 Extracted Lump Mass Model for 1999 Dodge Intrepid 11 Figure 4 2 Acceleration time history for 1999 Dodge Intrepid during frontal barrier crashes at 25 30 and 35 mph impact speeds 11 Figure 4 3 Time required for a 10 km hour change in velocity during a crash under various circumstances 12 Figure 5 1 MICFOCOBITOIIGT s 15 IIgure 5 2 Inmble AGCE I GPS UNN o E nei UR 17 Figure 5 3 Novatel CDPD wireless modem 19 Figure 5 4 Completed ACN Rev A 22 Figure 5 5 Basic Schematic of ANJEL Mobile Unit Rev A 23 Figure 6 1 An NMEA 0183 sentence 20 Figure 6 2 Automated Crash Notification via Wireless Web 30 Figure 6 3 Base Station Research 31 Figure 6 4 Sample Base Station Display 32 Figure 7 1 Micro drop Tower Apparatus 35 Figure 7 2 Drop tower results 3 Figure
4. TCR import ChER 0100917 1 INT VEC PRTO VEC ISR readAD interrupt ISR readAD ER inport TMDROL 66 status 0 0100 4 fullprog mod c ACN Rowan University this program is the first complete alpha prototype program for ACN 21 writes 20 reads VOI SERIO LIP ECI void crashtransmission define IBAUD 19200 1200 baud rate Wich modem either 24200 ot 1200 without modem gt 19200 9600 4800 etc define TBUFSIZE 384 On buffer define RBUFSIZE 384 Jus OF 4 19 date no parity 1 char 0 we don t want modem char ECHO 1 we do want character echo define CS4 0x40C0 define CS5 0x4100 Factine ticks 921 77 0198 432mhz720 0 001 cher Buf RBUPSIZESL 2 dummy buffer for receiving a complete command Loic Status main unsigned int input upper lower 0c 12 7 char tbuf TBUFSIZE IJ transmit char 74 buffet rnt ocrrcbutter 509 int index char Coordinates RBUFSIZE L status 0 0000 runningtotal 0 index 0 allows serial port 0 to work i ROM reload vee l4 DZ0 orrc 67 endif Interrupt PERTO initialize serial ports Dinit z1
5. Wireless Communication Subsystem Communication Technology Figure 5 3 Novatel CDPD wireless modem There are a number of possible wireless technologies that can be used for the transmission of the vehicle s location These technologies include Radio Frequency RF cellular and Cellular Digital Packet Data CDPD modems among others Rev A uses a CDPD modem manufactured by Novatel Wireless It is 0 6 W full duplex wireless modem It supports maximum transfer rates of up to 19 200 bps and uses a mere 8 mA in sleep mode This is important because in crash if the ACN unit is operating on backup batteries the system should use as little power as possible While this approach seems sufficient other possible communication means are possible and should be considered for a production system An important issue in determining the technology to use will be the available coverage for the given technology versus related cost While CDPD performs satisfactorily it would require consumers to purchase a monthly plan in order to use the ACN Cellular seems to be advantageous in this sense because with the new E911 standards being enforced any carrier that detects a 911 call must accept it This would save the consumer the cost of having to purchase a monthly plan of some sort with a cellular provider in order to use the crash notification system 19 The cellular antenna is of even greater concern than the GPS antenna
6. Figure 6 4 Sample Base Station Display Normally the APRS SA client software receives GPS strings from an APRS server via the TCP IP protocol In setting up APRS SA the user is given the option to select their APRS Server of choice For the prototype Base Station our approach was to develop an APRS Server look a like that received messages from the Mobile Units from a UDP port and served those messages to the APRS SA client from a TCP IP port The messages which were passed to APRS SA were formatted by our program to look like messages which would normally be received via packet radio This approach allowed the APRS SA program to believe that it was receiving packet radio messages when in actually it was receiving messages from a Mobile Unit 32 The UDP protocol was selected for the wireless communications link between the Mobile Unit and the Base Station instead of the more typical TCP IP The UDP protocol does not require verification of the transmitted message packets and hence is a faster protocol than TCP IP This approach removes the computational burden of verification on the limited computing resources of the Mobile Unit and allows the Mobile Unit to transmit repeatedly to the Base Station without having to pause after each transmission and wait for an acknowledgement The Base Station software was initially written as a Perl script and later rewritten as a Java application The TCP IP port was set as 9110 and the UDP port was
7. Project Manager Edward Kondrath NJDOT 14 Sponsoring Agency Code 16 Abstract The report describes the development of a Low Cost Automated Crash Notification system for eventual field testing on New Jersey highways The project was developed in response to national studies which show that nearly half of all traffic crash fatalities occur before the crash victim reaches a trauma center Many of these deaths can be attributed to the inability of EMS personnel to locate and reach the victim during the so called Golden after the accident when emergency medical treatment is most effective The goal of this project was to dramatically reduce EMS response time by developing and testing an advanced in vehicle system which automatically transmits the location and severity of a crash to EMS personnel Specifically the project has designed developed and tested a low cost functional system that combines wireless communications and Global Positioning systems with a network of inexpensive sensors for crash detection 17 Key Word 18 Distribution Statement Car Crash Emergency Medical Services Wireless Communications Automated Crash Notification 19 Security Classif of this report 20 Security Classif of this 21 No of Pages 1700 7 8 72 Reproduction of completed page authorized ACKNOWLEDGMENTS The authors wish to acknowledge William Hoffman Edward Kondrath Steven Kook and
8. 7 3 1999 Dodge Intrepid crash pulses in Full frontal barrier collisions at 25 39 MPN Impact 37 Figure 7 4 Concept Model of the Benchscale Impactor 38 Figure 7 5 Air cylinder with 3 bore diameter 40 Figure 7 6 Test stand with mounted piston 40 Figure 7 7 Rails mounted to test stand 2 2 2 41 Figure 7 8 Carriage on rail 1 41 Figure 7 9 Accelerometer mounted to testing plate 43 Figure 7 10 Schematic for data 43 Figure 7 11 Acceleration Pulses without the Mobile Unit Enclosure 45 Figure 7 12 Acceleration Pulses with the Mobile Unit Enclosure 45 Figure 7 13 Acceleration Pulse of the Air 46 LIST OF TABLES Table Sal Charging etu sd E ut n pus Table 5 2 Cost for ANJEL Mobile Unit Rev Table 7 1 List of Mini Sled Tests 1 SUMMARY report describes the development Low Cost Automated Crash Notification System for eventual
9. Nicholas Vitillo of the New Jersey Department of Transportation for their support of this research effort The authors also wish to gratefully acknowledge the invaluable efforts of the following Rowan University undergraduate engineering research assistants who have made the success of this project possible David Browning Aditya Chaubal Peter Ferrara Michael Gilligan Samuel Greenfeld Devon Lefler Amol Shah and Amip Shah 5 1 2 Introduction and Background 2 3 System Requirements Architecture 4 4 9 5 Mobile Unit System Description 13 6 Base Station System Description 27 VOSUMG Ve Sicca 34 0 48 9 a aiai 49 10 gcc 50 Appendix Source Code for the Base Station Prototype 51 Appendix SISAME Model of a Dodge Intrepid 58 Appendix C Source Code for the Mobile Unit Prototype 60 LIST OF FIGURES Figure 2 1 The Objective of Automated Crash Notification is to Improve Emergency Response
10. Or HP e rR HH CO 0 POLCAP SMT 10UF CAP SMT 20UF CAP SMT 33UF CAP SMT 100UF 5400 5500 POINT JUMPER CM7000 26PIN HEADER 26PIN CONN 931233 91 8 568 93F1233 906 3174 EG1957 226 1011 APPP 001 EG1500 C20 039 222 C48 COT U27 eno C36 C31 C42 C34 C49 C50 C33 C237 C46 R6 R7 R8 R1 TP 7 10 13 16 19 EOD 5 8 11 14 H3 H4 5 U1 H1 C3 5y C37 C43 C45 C41 47 6 9 2 BI 8 gt PO CAP100RP SMT B 5 5 D 5 5400 5500 POTNIT IDC8M 2MM CONN40M IDC26F 25 1 uF 10 uF 20 uF 33 uF 100 uF 10k 400 1 uF 16V Tantalum 10uF 16V Alum Elec SMT 20uF 16V Alum Elec SMT 33uF 16V Alum Elec SMT 100uF 35V Alum Elec SMT Resistor 1 4W Resistor 1 4W Test Points 2mm 8 pin Header 40 Pin Connector for CM7200 26 Pin Header for External Box 26 Pin Connector for External Box Receptacle for outer box Plug for outer box Cable clamp for outer box Heat Sink 220 Backup battery 1 2 V 600 mAHr SUB TOTAL FOR ACN UNIT Square post receptacle 3 pos switch Switch knob Cigarette Adapter plug Rocker switch Power SUB TOTAL FOR DEV UNIT TOTAL COS
11. field testing on New Jersey highways The system was developed in response to national studies which show that nearly half of all traffic crash fatalities occur before the crash victim reaches a trauma center Many of these deaths can be attributed to the inability of EMS personnel to locate and reach the victim during the so called Golden Hour after the accident when emergency medical treatment is most effective The goal of this project was to dramatically reduce EMS response time by developing and testing an advanced in vehicle system that automatically transmits the location and severity of a crash to EMS personnel Specifically the project has designed developed and tested a low cost functional system that combines wireless communications and Global Positioning Systems with a network of inexpensive sensors for crash detection 2 INTRODUCTION AND BACKGROUND Figure 2 1 The Objective of Automated Crash Notification is to Improve Emergency Response Times With the advent of trauma centers the fatality rate of persons reaching a hospital after a car crash has dropped dramatically over the last twenty years However nearly 20 000 crash victims die every year before ever reaching the hospital NHTSA 1999 Undoubtedly some fraction of these deaths result from catastrophic crashes However many of these deaths can be attributed to the failure of EMS personnel to reach the victim during the so called Golden Hour after the accident
12. for if this antenna is lost no transmissions will be possible Multiple antennas for the cellular unit possibly in the front and the back of the car would be advantageous as this would give maximum antenna survivability in a crash However there are other issues related to the number and location of antennas Wires must be run from each antenna to the crash notification box and an excessive number of antennas would intensify the associated labor thereby reducing the ease of installation of the system Moreover aesthetics is also a very important issue The presence of antennas in safe but obscure locations might actually have an adverse affect on the marketability of the product as far as the consumer is concerned More studies need to be conducted in this area to determine the consumer s preferences Power Requirements Rev B The ACN power system consists of a power conditioning system of the various filters and regulators required to convert the 12V DC from the car battery into the necessary DC voltages vital to the internal circuitry The power system also includes a back up battery pack complete with its own charger and a mechanism for switching between primary and back up power consumption modes All of the circuitry relies on 5V DC with the exception of the CDPD modem for which 3 6V DC must be supplied In addition a 5V analog reference is needed for the A D converter onboard the microcontroller To provide each of these required vol
13. obtained of the deceleration time history of the carriage the impact point at the end of the rails The first two graphs were taken without the Mobile Unit Enclosure Figure 7 11 The same tests were performed twice first with the Teflon padding then without These show that only about 8 5 total acceleration was obtained Note that the use of Teflon pads produced only moderate increases in peak acceleration The procedure was then repeated with the ACN box figure 7 12 Figure 7 12 shows that similar results were obtained In all cases peak deceleration was below 10 G 0 0 0 005 0 01 0 015 0 02 0 025 0 03 0 035 0 04 Time s Figure 7 13 Acceleration Pulse of the Air Piston In addition to these impact tests performed on the rails tests were also performed on the air cylinder piston Figure 7 13 shows that the peak acceleration of the air cylinder piston was about 35 0 5 The peak was attained early in the test when air cylinder pressure was at its highest As the piston shaft extended the acceleration dropped back to zero as the cylinder air pressure dropped The entire event was observed to take place in approximately 35 milliseconds After tests on the Mobile Unit enclosure itself were completed a test was performed with the Rev A Mobile Unit in place When the Mobile Unit was plugged into the and slammed into the wall the word CRASH appeared on the screen when appropriate indicating the success of the system i
14. the acceleration values from the accelerometer will be logged by the micro controller which will then integrate these values over a 40 ms time interval to determine the change in velocity of the vehicle This change in velocity is then compared to a predetermined threshold which allows the microcontroller to determine whether or not a crash has occurred v Firewall Radiator Occupant Compartme Figure 4 1 Extracted Lump Mass Model for a 1999 Dodge Intrepid Acceleration g s 20 25 30 35 40 0 000 0 025 0 050 0 075 0 100 0 125 0 150 Time sec Figure 4 2 Acceleration time history for 1999 Dodge Intrepid during a frontal barrier crashes at 25 30 and 35 mph impact speeds 11 Percentage Impacts That Experience 10km h Change in Velocity at or Below a Time Value Percentage O 27 29 29 30 31 32 33 34 35 36 5 Figure 4 3 Time required for 10 km hour change velocity during crash under various circumstances To determine the different possible threshold values for a crash numerous crashes were simulated using the SISAME impact simulation code developed by NHTSA Mentzer 1999 First a model of a 1999 Dodge Intrepid was extracted using crash test data available from the NHTSA Vehicle Crash Test Database The detailed SISAME model file is provided as a
15. the value read to the serial port for reading with the hyper terminal define IBAUDO 4800 1200 baud rate define IBAUD1 9600 1200 with modem either 2400 or 1200 without modem gt 19200 9600 4800 etc define TBUFSIZE 384 size of transmit buffer define RBUFSIZE 384 size of receive buffer define CS4 40 define CS5 0x4100 char MODE 4 8 data MO parity tob char MODEM 0 we don t want modem char ECHO 1 we do want character echo main unsigned int input upper lower i THERE ALIS IDE GOUNE char tpu TBUFSIZE transmit buffer char rbuf RBUFSIZE Lf receive butter char buf RBUFSIZE 1 dummy buffer for receiving a complete command bDutzlkBUESIAZETL s char output char 16 allows serial port 0 to work 1 reload vecTtl4 DzO corro xn endif Dinit z0 rbuf tbuf RBUFSIZE TBUFSIZE MODE IBAUDO NO MODEM ECHO 222 Dinit z1 rbuf tbuf RBUFSIZE TBUFSIZE IBAUD1 NO MODEM 35 reference 0 0 reference 1 1 reference 2 2 60 reference 3 3 reference 4 4 perterence 5 157 reference 6 6 reference 7 7 reference 8 8 reference 9 reference 10 pererence lr B7 reference 12 C pererence lis DD reference 14 E reference 15 input 0 endless loop constantly monitoring for new com
16. when the Z1 port is 74 anitralized Dinit z0 rbuf tbuf RBUFSIZE TBUFSIZE MODE IBAUDO NO MODEM ECHO Dinit 21 tbuf RBUFSIZE TBUFSIZE MODE IBAUD1 NO MODEM ECHO endless loop constantly monitoring for new command line runwatch if Dread zO buf ENTER 0 Le wait LOE string terminated with CR Dwrice cxlo bury strleni bour 272 62 3 crastest cpp physical crash test code ACN Rowan University define sets named constants define IBAUD 19200 1200 baud rate with modem either 2400 or 1200 without modem gt 19200 9600 4800 etc define TBUFSIZE 384 Of transmit define RBUFSIZE 384 gize rfecerve butter char MODE 4 g Cate Stop char NO MODEM 0 we don t want modem char ECHO 1 we do want character echo fdefine CS4 0x40C0 define CS5 0x4100 define ticks 461 19024432 20 10001 define ticks 921 18 432 MHz 20 0 001 VOLA PREO Cae o int status main unsigned int input upper lower ghar tout TBUESTZE 4 transmit buffer char rbuf RBUFSIZE buffer ine 50 int int index int Crash IINE i gt declare an array coordinate of size 1 greater than buffer char coordinate RBUFSIZE 41 Le int maxruntotal tnt minr ntotaLl mE MAr rUn Otal Dk int SGEEnguntotel Dk char o
17. 1 666 5682 666 5682 666 5682 58 51 8 050754 SpriD Firewall Descr Firewall NegMass OccComp PosMass Engine StaType SI SU 877 2985 51 0 X 0 21 78571 108 9286 130 7143 217 8571 239 6429 F 0 0 0 0 768 7556 19880 03 51 0 MMax 1 SpriID Occ Wheels NegMass OccComp PosMas Descr Occ Wheels s Wheels ST 1066 072 StaType SI SU 230 3747 X 0 7 276681 36 38341 43 66009 72 76681 80 04349 F 0 0 0 0 0 0 51 0 MMax 1 15 NegMass Wheels PosMass StaType SI SU 62 11453 0 44 25638 88 51277 F 0 0 794 2044 51 1 4172 Output Information OutClass MassTS Comments Descr Wheels Eng 5 293 0707 8 851277 53 10766 97 36405 0 0 1343 997 MMax 3863 411 51 92 60249 43 57143 T5259 261 4286 0 0 38992 6 XS1k 5 12806 14 55336 9504993677 87 32017 0 0 0 XSlk 21 big 0259 61 95894 106 2153 0 0 1632 624 69 420 6391 Qty AVD Mass 59 9767 65 35714 174 2857 283 2143 0 0 58105 17 21 83004 96 41345 94 59685 0 0 0 2 OLIO 70 81022 115 0666 0 0 1632 624 87 14286 196 0714 305 0 0 76637 24 29 10672 65 49013 1018755 0 0 0 35 40511 79 66149 L23979 0 244 4115 1632 624 APPENDIX SOURCE CODE FOR THE MOBILE UNIT PROTOTYPE 1 A_DACN cpp This program tests the A D and outputs
18. 9111 The code for both versions is provided as an appendix to this report Future Work The long term objective of the ACN system is to connect the Mobile Units with existing or expanded 911 systems However this effort will require coordination with existing 911 system operators and careful attention to how best to present crash information graphically to operators who are more accustomed to receiving voice only calls The Base Station developed here will provide an early evaluation of possible 911 operator user interfaces The Base Station may also be suitable for limited field testing of the system for captive fleets such as the State Police or NJDOT vehicles 33 7 TESTING evaluate the performance of the ANJEL system the Mobile Unit was subjected to a battery of tests during development The tests included both non impact vehicle tracking test as well as low severity impact tests This section describes the test strategy test procedures test apparatus and test results Tracking Test To check the communication between the Mobile Unit and the Base Station the completed prototype was tested in tracking mode In this test the Mobile Unit and associated antennas were mounted in a car and the Mobile Unit was switched to its special diagnostic tracking mode When in tracking mode the Mobile Unit automatically reads the GPS and transmits its location every second Note that tracking mode is a research diagnostic only this mode w
19. AirMed March April 1998 5 Preziotti G Kanianthra J and Carter A Enhancing Post Crash Vehicle Safety through Automatic Collision Notification Proceedings of the 17 International Technical Conference on the Enhanced Safety of Vehicles Amsterdam June 2001 6 Mentzer S Sisame User s Manual National Highway Traffic Safety Administration 1999 7 Trimble Navigation Limited ACE II GPS System Designer Manual June 1999 8 Conexant Systems Zodiac GPS Receiver Family Designers Guide February 1999 50 SOURCE CODE FOR THE BASE STATION PROTOTYPE A 1 Base Station in Java Function The following stand alone Java server received UDP messages from the Mobile Unit and served these messages from a TCP IP socket to an APRS SA client import java net import java io public class Server public final static int MAX PACKET 514 65507 public static void main String args oystem out println Base Station Server n String hostname localhost int port 8080 byte buffer new byte MAX_PACKET_ SIZE 1 Setup Datagram Socket try DatagramSocket soc new DatagramSocket port DatagramPacket thePacket new DatagramPacket buffer buffer length Read keyboard messages and send them to the UDP Socket BufferedReader in new BufferedReader new InputStreamReader System in while true soc receive thePacket 51 String new Stri
20. FHWA NJ 2001 027 DEVELOPMENT OF A LOW COST AUTOMATED CRASH NOTIFICATION SYSTEM Final Report July 2001 Submitted by Dr H Clay Gabler Associate Professor Rowan University Department of Mechanical Engineering Glassboro NJ 08028 NJDOT Research Project Manager Edward Kondrath In cooperation with New Jersey Department of Transportation Division of Research and Technology Trenton NJ 08625 DISCLAIMER STATEMENT The contents of this report reflect the views of the authors who are responsible for the facts and the accuracy of the data presented herein The contents do not necessarily reflect the official views or policies of the New Jersey Department of Transportation or the Federal Highway Administration This report does not constitute a standard specification or regulation Technical Report Documentation 1 Report No 2 Government Accession No 3 Recipient s Catalog No AAO ee 4 Title and Subtitle 5 Report Date Development of a Low Cost Automated Crash Notification System 7 Author s 8 Performing Organization 9 Performing Organization Name and Address 10 Work Unit No TRAIS Rowan Universit Department of Mechanical Engineering 11 Contract or Grant No Glassboro NJ 08028 12 Sponsoring Agency Name and Address 13 Type of Report and Period New Jersey Department of Transportation voveled Division of Research and Technology P O Box 600 Trenton NJ 08625 0600 15 Supplementary Notes
21. Modem Antenna Cable T999 1 GPS U3 IDC8M 2MM GPS Unit 160 00 1 MAGMOUNT ANTENNA GPS Unit Antenna 45 00 1 74HCTO2 U9 DIP14 Quad 2 Input NOR 0 39 1 74HCT32 17153 DIP14 Quad 2 Input OR 0 41 2 LMC662 U22 U25 DIP8 Dual Operational OpAmp oe 1 LP339NA U19 DIP14 Ultra Low Power Dual Comparator 1 40 1 MAX232 U2 DIP16 RS 232 Transciever BO 1 74HCT259 012 DIP16 8 Bit Addressable Latch 0 85 1 74HCT374 U8 DIP20 Octal D Type Register 0 83 1 74HCT541 U24 DIP20 Octal Buffer Line Driver 2 T4LVX4245 U20 U23 SOIC 24 3 3V 5V Level Shifter 3 34 1 V33ZA2 U30 V33ZA2 MOV 26 V 0 54 1 171086 9445 220 3 6V 1 5A Low Regulator Suc 1 LM2940 U16 220 5V 1A Low Regulator 2 45 1 NJM7809 117 220 9V 1 5A Regulator 0 63 1 114040 014 TO 92 5V Precision Reference 3 47 3 EMI FILTER B2 L3 SMT 10000pF 10000pf 50V EMI Filter 2 10 1 IND 15 400 275 uH High Current Toroid 1 1 5404 D1 267 03 400V 3A Silicon Rectifier 05205 1 ADCO809CCNA 04 DIP28 8 Bit uP Compatible A D 6 90 1 ADXL250 SOP 07 SOIC 14 Dual Axis Accelerometer 23 94 1 5 04 U5 DIP16 Sample amp Hold 8 15 1 LM34 U21 TO 92 Temperature Sensor 8 44 1 MX7528 U26 DIP20 DAC Unit 7 11 1 45 031 DIP8 Crystal Oscillator 1 MHz 2 78 2 CAP100 C25 NS CAP100 10 pF 10 pF 100V Ceramic Capacitor 0 38 29 CAP200 C4 CAP200 0 1 uF 0 1 uF 50V Ceramic Capacitor 3 48 COn CT Cs CES CI C21 C24 C257 C26 24
22. T FOR ACN PROTOTYPE 62 72 42 80 18 06 3 63 1 66 2 07 30 40 58 64 6 42 18 00 786 54 07 255 225 78 PNM A OO 24 54 811 08 ADDITIONAL COSTS INCLUDE PCB FABRICATION RAW MATERIAL AND OTHER MECHANICAL COSTS WASHERS SCREWS ETC 26 6 BASE STATION SYSTEM DESCRIPTION In the event of a crash the Mobile Unit will automatically notify the Base Station of the crash via a wireless communications link The functions of the Base Station system are to 1 receive the simulated emergency call from the Mobile Unit 2 retrieve GPS data and crash severity as transmitted by the Mobile Unit and 3 display the location and severity of the simulated crash using computerized maps for Emergency Response Team dispatch Design concerns include how best to present crash location and severity to the Base Station operators and how to ensure that large numbers of calls can be handled simultaneously The discussion below will detail the Crash Notification Message Content approaches for Crash location mapping the wireless web communication strategy and the software implementation Message Content After detecting a crash the Mobile Unit must transmit a message to the Base Station which describes the crash location and severity Knowledge of the crash location allows the EMS center to dispatch EMS crews to rescue the crash victim Knowledge of the crash severity provides
23. YSTEM j APRS BELAYSChrlSSchrlo Snewip Snewconn gt peerhost Sport Snewconn gt peerport Stime scalar localtime Slog I NFO TCP Connection from porc Sport Add new Connection to selector 54 Sselector add newconn elsif Sconnection Sudp UDP connection possible alert v connectron recv udpdata 220 Code to verify valid alert goes here my lines split n Sudpdata for my Sline Glines 4 Sline s r go system echot Jot Tor Line ot lanes if Sudpdata NSNWNWGGA pd NO Xd 10 Begin if valid GPS data Shour 51 Smin 2 Ssec 53 Slat 5422 548 925 Slong 56 Sew S7 Squality 8 Sns uc Sns Sew uc Sew my Sudp gt peerhost my Sport Sudp gt peerport amp log ALERT Alert Recieved Stime from Sip port Sport amp log ALERT Sip GPS Data 1 SlongSew amp log ALERT Sip GPS Data Recorded Shour smin sec GMT ALERT GPS Quality indicator Code to write to MS s stuff goes here print NMEA Sudpdata Pf Sudpdate gt forint NMEA open gt Scrashfile amp log WARNING unable to trigger MS crash system via file print FLE 242 close Code to alert all via APRS goe
24. a common channel shared with other users to the destination computer which reassembles the message The result is a continuous Web connection between the Mobile Unit and the Base Station which avoids the dial up delays which are inherent in circuit switched designs Unlike the circuit switched design which has the potential for phone call contention problems the number of accidents which can be handled by a Web based ACN Base Station is general limited primarily by the bandwidth of the Base Station Internet connection ACN Server Base Station CDPD Wireless ACN Clients Mobile Units Figure 6 2 Automated Crash Notification via Wireless Web 30 Base Station A Research Prototype Base Station was developed which implements the functional requirements described above As shown in Figure 6 3 the Research Prototype consisted of a Dell Dimension 600 MHz Pentium III running Windows 98 equipped with a high speed Internet connection In the event of a crash the Mobile Unit and Base Station will communicate using wireless Cellular Digital Packet Data CDPD technology over analog cellular networks CDPD is a new wireless Web access technology with widespread coverage in the eastern United States CDPD allows a direct TCP IP link to be established between the Mobile Unit and Base Station Using CDPD the Base Station is designed as a Web Server and the Mobile Unit reports a crash to the Server via a wireless Internet conn
25. a crash the microcontroller will have the position of the car to within a second before the crash For testing purposes in Rev A the GPS output is read into a computer using RS 232 However as the GPS actually outputs TTL logic levels once the GPS unit is embedded into the system no interface will be required between the GPS unit and the microprocessor The C code for this operation is shown in the file CDPD GPS cpp included in Appendix C at the end of this report Antennas A major concern related to the usage of GPS is its antenna While special care can be taken to ensure to crashworthiness of the antenna there is always the possibility that the antenna may be destroyed in a crash thereby rendering the system useless To combat any such issues it was decided that the GPS unit should automatically update the vehicle location every second As a result even if the GPS antenna is lost in a crash the system will still be able to transmit the last known position which given the sample rate will be to within a second before the crash Note that two antennas are required for the system one fore the GPS and one for CDPD communication For Rev A the two antennas received from the GPS and CDPD vendors were used in their modified form achieve lower costs Conexant Zodiac GPS receiver was used Rev The Zodiac receiver provides performance similar to the ACE II receiver a comparably sized package
26. ata my aprssystems Sselector gt can write 1 for my Swrcon aprssystems if 5 Slistener amp amp Sudp amp log WARNING Notifying APRS at Swrcon gt peerhost Sudpdata lt 5 13 101 00 prrut ochr 4 End End For End else wierd string else For TCP connections Sconnection gt recv Sa 1 if Sa ne undef Got one in Ssdonothing 1 End if not equal to the undefined value else Close connection my Stemp connection peeraddr my Sconnection gt peerhost Shere Sselector count 2 3 Slog INFOS Closing 2 port econnmectrron opeerpoft s Sselector gt remove Sconnection close S connection Clog INFOS gone Slett APRS client Ss remain End else close connection End else character or close 4 End for 56 End while consready print nI exited wrong timeout waiting for users maybe n close LOGGER close NMEA sub log Stime scalar localtime print dE cp prine LOGGER Stimer d mt 57 APPENDIX SISAME MODEL OF DODGE INTREPID The following SISAME model was used to develop the crash detection algorithm for the Mobile Unit SISAME Input File Run Information RunID INTREPIDFF Title INTREPID 1999 Full Frontal Model Weight Extraction DimSys Metric De
27. cations and Global Positioning systems with a network of inexpensive sensors for crash detection The project has developed two Mobile Unit prototypes which have been demonstrated to communicate vehicle location to a remote Base Station a Wireless Web communications link e The project has developed a Base Station based upon commercially available mapping software which has been successfully demonstrated to communicate via wireless modem with Mobile Units in the field receive vehicle coordinates from the Mobile Unit and automatically indicate Mobile Unit location on the Base Station system e Two impactors were developed for testing of the Mobile Unit a a Drop Tower and b a Pneumatic Benchscale Impactor Of the two the Drop Tower was found to be the more promising The Drop Tower was found capable of producing up to 100 G s and produced crash pulse shapes which were observed to be similar to actual rigid barrier crash tests with production passenger cars e n low severity impact tests the research team has successfully tested the Mobile Unit at severities up to 9 G These tesst were designed to evaluate the survivability of the electronics to impact as well as testing the ability of the system to detect and report collisions of this magnitude 48 9 5 This project has successfully demonstrated the feasibility of a Low Cost Automated Crash Notification System The performance of both a Mobile Unit pr
28. ch could a detect a crash b determine crash location and c transmit crash severity and location to a Base Station These were the primary design objectives for the first prototype i e to demonstrate functionality Although other design considerations e g cost size power requirements ease of installation and crash survivability would be important in the eventual production Mobile Unit these design criteria were relaxed during pursuit of the first prototype To facilitate demonstration of proof of concept and retain maximum design flexibility Rev A was envisioned as a research prototype Rev A was intended to serve as a test bed for potential ACN technologies including crash sensors GPS chip sets and wireless communication components Rev A was designed to be as modular as possible so that alternate components e g GPS boards could be readily swapped into and out of the prototype Mobile Unit to investigate improved performance Rev A was also designed with numerous internal diagnostics to track and allow debugging of system performance during operation in the field Finally because this was to be a research prototype cosmetic packaging concerns were postponed until the development of later prototypes This allowed antennas for example to be placed where convenient for testing as opposed to attachment points more aesthetically pleasing to a consumer Similarly this approach allowed power for the Mobile Unit to be obtai
29. ch crash sensor In these future systems the Base Station operator will be presented with a display not only of where the collision took place but also with a separate display which shows the crash severity Knowing the crash severity the operator will then have an early warning of the expected level of injuries at the crash site Other Information The message may also contain supplemental information to better identify the car to EMS personnel Fields such as the car VIN make model model year and car color should be considered for future systems It should be noted that many of these fields can be determined from the VIN If VIN is available from the Mobile Unit message future systems may be able to tie into state Vehicle Registration databases to identify the owner of the vehicle to expedite notification of family members Crash Location Mapping Upon receipt of an emergency message from the field the Base Station will present a map to the operator showing the location of the crash site Numerous commercial GIS mapping products e g ArcView exist for providing this function However these packages tend to be relatively expensive Asa less expensive alternative several consumer mapping products were investigated for their ability to provide this function None of these packages are of course designed for Automated Crash Notification However they do provide a database of street level maps for integration into a Base Station softwar
30. d to survive crash loadings typically 30 G in a 35 mph crash and potential of power after the crash Antennas for GPS and wireless transmission must survive the crash so that the crash location can be determined and notification of the crash event can be transmitted to the Base Station As not all transmissions between the Mobile and Base Unit may be received the Mobile Unit must be designed to transmit multiple times Crash survivability can be increased through several means including 1 backup battery power 2 locating the Mobile Unit inside the occupant compartment cage 3 taking GPS measurements repeatedly during normal driving and transmitting the last known location to the Base Station if the GPS lock is lost The post crash operation of the system was evaluated in laboratory testing at Rowan University Base Station Functional Requirements The Base Station system must 1 receive the simulated emergency call over the Mobile Unit 2 receive GPS data and crash severity from the simulated crash site and 3 display the location and severity of the simulated crash using computerized maps for Emergency Response Team dispatch Design concerns include how to best present crash location and severity to the Base Station operators and how to ensure that large numbers of calls can be handled simultaneously The long term objective of the ACN system which will not be conducted under this research effort will connect the Mobile Units
31. e package was written especially for this project Of these consumer products the most promising programs for the Base Station application were Street Atlas Version 8 0 by DeLorme and Mappoint 2002 by Microsoft Both products provided the street level detail required for the Base Station operator to direct EMS teams to a crash site and both products were capable of being controlled by an external program Street Atlas is the mapping product used by the APRS SA shareware amateur packet radio location server Mappoint 2002 provides a suite of Active X controls which allow external program access to mapping display functions 29 Wireless Communication Subsystem Design One key enhancement of this system over existing ACN concepts is Mobile Unit to Base Station communication over the wireless web Existing ACN systems are typically based upon circuit switched communication in which the wireless network assigns a dedicated frequency to the call between the car and the Base Station There are only a limited number of these frequencies When they are expended as many mobile phone users have experienced the result is that phone calls do not connect In the Rowan system on the other hand each car has a unique IP address and wireless communication is conducted using packet switching as shown in Figure 6 2 In packet switching the signal is divided up into individual packets of data tagged with the address of the destination and transmitted over
32. e the typical frontal barrier crash has a duration of approximately 150 milliseconds while panic braking requires over 1000 milliseconds Message Content When crash is detected the Mobile Unit must transmit a message to the Base Station which describes the crash location and severity Knowledge of the crash location allows the EMS center to dispatch EMS crews to rescue the crash victim Knowledge of the crash severity provides the EMS center with an early snapshot of the seriousness and potential injury consequences of the accident The message to the Base Station must include both these data facets as well as information detailing the time of the crash and a description of the car Crash location can be as straightforward as the GPS location longitude and latitude Crash severity will be provided for each crash sensor and will be either the change in velocity or the crash pulse along each axis It should be noted that while the crash pulse requires transmission of a longer message the crash pulse typically provides sufficient information to infer whether the car struck a tree or another car which may require additional EMS personnel Inclusion of crash severity for each axis allows the Base Station to distinguish between frontal and the potentially more serious side impacts Crash Survivability The Mobile Unit must be capable of surviving and properly functioning after a crash The unit its enclosure and necessary antennas must be designe
33. e experimental procedure was drop the plate with accelerometer from various heights on to different surfaces at the base The plate was dropped from three different heights six feet eleven feet and seventeen feet Three different densities of foam ranging from very soft to very dense were used for the impact surface A total of eighteen drops were performed To evaluate test repeatability two drops were performed for each height onto each of the foams During testing the accelerometer plate was pulled up to the required height and the data acquisition system was started The plate was then allowed to freefall onto the foam The acceleration was recorded for each drop and plotted versus time Results As shown in Figure 7 2 the drop tower was able to produce realistic accelerations when tested with the Mobile Unit enclosure However while the absolute acceleration levels were realistic the results of the test were not entirely representative of an actual crash For example Figure 7 3 shows the results of a 1999 Dodge Intrepid frontal crash test Comparison of the two types of tests shows that while the acceleration levels are similar the pulse shapes are quite different This would suggest that the way the forces are applied to the test articles differs in some manner Moreover in the drop test the impact speed 15 limited to a terminal velocity of approximately 20 miles per hour which is at the lower range of real world injury producin
34. ection This approach allows the Base Station to monitor multiple vehicles involved in crashes without the requirement for banks of dedicated phone lines When the Base Station receives a message from a Mobile Unit the Base Station displays the crash location and severity on a commercially available mapping product Figure 6 3 Base Station Research Prototype The system will use Cellular Digital Packet Data CDPD sometimes referred to as a Wireless IP connection to transmit data between the Mobile Unit and the Base Station CDPD is a cutting edge wireless communications protocol which 31 allows direct connection of remote devices to the Internet In addition to CDPD the Mobile Unit has been designed for adaptation to other wireless communications options including CDMA Code Division Multiple Access Data GSM Global System for Mobile Communications and emerging third generation wireless protocols e g GPRS General Packet Radio Service and W CDMA Wideband Code Division Multiple Access Software Implementation The Research Prototype Base Station was implemented using the APRS SA Packet Radio Location Software APRS SA is a shareware software package which automatically plots the location of a transmitted GPS string on maps displayed under Street Atlas 8 0 Figure 6 4 shows a map displayed by the Base Station running APRS SA during a tracking test of the Mobile Unit near Rowan University Fowan College OF Jersey
35. ed to accomplish this task In choosing an air cylinder there were two selection factors the bore size of the cylinder and the stroke length of the piston In the current design the stroke length was fixed at 12 in and the bore 38 size was adjusted accordingly The major factor affecting the bore size would the force to accelerate the load of the carriage Using 5 second law F ma with an assumed Mobile Unit mass of 710 and an acceleration of 12454 28in s 30 G s the required force was found to be 225 6 Ib In other words the air cylinder must be able to produce a force of at least 225 The next task was accounting for any losses in the system The work requirements for the pneumatic cylinder were computed as follows m 71 546 72 5 W Fx 1 2mV W 2707 4531 in To account for frictional and other losses in the system a cylinder work effiency of 50 was assumed The process was then re analyzed assuming polytropic work k 1 4 2 2a Z 2 Ve Bad 110 psia W 22 ee 2 2 1 0 110psia 27 p B 2707 453 b in _ 2 1 1 4 2 475 However in addition to accelerating the carriage up to 30g and overcoming frictional losses inside the piston the cylinder must also ensure that the air can be evacuated prior to the next stroke any backflow would prevent the proces
36. ency Response Systems One of the most challenging and important questions confronting deployment of an Automated Crash Notification system is determining how to integrate an ACN system into existing 9 1 1 systems Follow on research should actively consult with the representatives of the emergency response community to address this issue Lower Cost Wireless Communication New 3 Generation Wireless communications protocols will be introduced to the market in the coming months which should be actively investigated in follow on studies Currently the Wireless Modem accounts for half the cost of the Mobile Unit These newer wireless protocols have the potential to tremendously wireless communication performance and further reduce the costs of the wireless link between the Mobile Unit and the Base Station 49 10 5 1 National Highway Traffic Safety Administration 1999 Fatality Analysis Reporting System U S Department of Transportation 1999 2 National Highway Traffic Safety Administration Traffic Safety Facts 1999 U S Department of Transportation 2000 3 Thomas S G Smart cars need fewer brains and more old fashioned common sense U S News amp World Report February 14 2000 4 Champion HR Augenstein JS Cushing B Digges KH Hunt R Larkin R Malliaris AC Sacco WJ and Siegel JH Automatic Crash Notification the Public Safety Component of the Intelligent Transportation System
37. g automotive crashes Consequently a second impactor was explored to attempt to produce more real world impact conditions 36 Figure 7 2 Drop tower results E EH EN ON Intrepid 25 0cc Intrepid 30 0cc Intrepid 35 0cc C 9p m 0 i 15 20 AN 0 000 0 025 0 050 0 075 0 100 0 125 0 150 25 30 35 ae ed 40 Time s Figure 7 3 1999 Dodge Intrepid crash pulses in Full frontal barrier collisions at 25 30 and 35 mph impact speeds 37 Mini Sled Impactor System To produce higher speed impacts a bench scale pneumatically driven sled was designed and evaluated for possible Mobile Unit testing To accelerate the Mobile Unit enclosure the casing was mounted on a carriage placed between two guiding rails as shown in Figure 7 4 A pneumatic cylinder was selected to generate sufficiently high forces to rapidly accelerate the Mobile Unit at target speeds up to 35 mph Once the ACN unit had been accelerated to sufficiently high speeds it was crashed into a stationary wall at the end of the slide The design and construction of the required apparatus is described below Figure 7 4 Concept Model of the Benchscale Impactor Air Cylinder In the design of the impactor the biggest challenge was discovering how to accelerate the Mobile Unit until it reached a speed of about 35 mph An air cylinder was us
38. g the location of the crash and communicating crash severity and crash site location to the Base Station Figure 3 1 presents the system architecture of the proposed device The system consists of a single chip embedded microcomputer which is connected to a crash sensor a Global Positioning System GPS sensor and an embedded wireless modem In the event of a crash the crash sensor s will detect the vehicle impact and output a signal proportional to the deceleration of the vehicle The crash sensor signal output will be continuously monitored by the microprocessor which will decide whether or not a crash has taken place Upon detecting a collision the microprocessor will poll the GPS sensor to determine the final resting position of the car The microprocessor will then use its wireless modem to establish a communications link with the Base Station Once a link has been established the Mobile Unit will transmit crash site location and the crash severity to the Base Station Ideally the entire process including linkup will be completed within 30 seconds after the crash occurred giving EMS personnel a crucial edge in rapidly reaching the crash victim The Mobile Unit will be installed either under the driver s seat or in another occupant compartment location Locating the Mobile Unit in the occupant compartment will provide an accurate measure of the deceleration experienced by the occupants in a crash and will protect the Mobile Unit with t
39. git This makes it read 37 degrees and 44 953 minutes The field directly following that indicates whether it is north or south of the equator in this case N is indicative of north The longitude fields behave in a similar way with the only major difference being that the separation comes after the third digit So for example the longitude in fig 6 1 reads 122 degrees and 25 319 minutes west of the Greenwich meridian Crash Severity One of the parameters most crucial to predicting crash victim injury level is crash severity Crash severity is a direct measure of the mechanical forces which lead to human injury The most important measure of impact severity is the crash acceleration deceleration time history frequently referred to simply as the crash pulse If the crash pulse is known both delta V and other impact severity measurements such as average acceleration level can be calculated Measurement of the crash pulse is a key instrumentation requirement of the majority of full systems laboratory crash tests 28 Crash severity is computed by the Mobile Unit by analysis of the crash pulse read by the onboard crash sensors It is this crash severity in fact which is evaluated to determine whether to initiate the emergency call from the Mobile Unit to the Base Station Initial tests of the Mobile Unit have included the crash location alone Future systems will include the delta V and or the crash pulse as read from ea
40. h can detect that an accident has taken place and automatically notify the emergency medical personnel of the severity and precise location of the accident Once activated an Automated Crash Notification system would automatically transmit a signal to a 9 1 1 dispatch center where an electronic map would pinpoint the signal location Precise location of the traveler in trouble enables rapid emergency response More advanced sensors can also estimate the injury producing capability of the crash The first estimates of the number of potential lives saved by ACN technology are 3000 lives per year Champion et al 1998 The National Highway Traffic Safety Administration has sponsored a trial ACN system Preziotti et al 2001 This program is in the process of installing ACN in 1 000 privately owned vehicles in upstate New York The ACN system uses on board sensors to identify that a crash has taken place It then uses the Global Positioning Satellite GPS system and conventional cellular phone systems to deliver a message based on the sensors input directly to 911 operators While promising this system has proven to be extremely expensive To date the total Federal cost of the study has been about 3 million The technical approach used in this project has resulted in an estimated 500 cost per unit motivating a search for a lower cost approach to Automated Crash Notification Objective The goal of this project is to develop and test an ad
41. he first z axis sensor by a known distance would allow detection of rollover The system uses a newly developed low cost crash sensor the Analog Devices ADXL 250 These crash sensors are inexpensive silicon based accelerometers which were initially developed for airbag systems and cost two orders of magnitude less than conventional accelerometers GPS Sensor The system uses a newly developed low cost GPS sensors the Trimble ACE II system and the Conexant Zodiac System These sensors provide location resolution under 30 meters Two options were investigated for GPS data processing for the mobile unit The first option was to use a turn key single board system which processes the raw GPS data on board to determine the position of the car The second option considered was to transmit the raw GPS data directly to the Base Station that will compute crash site location using a more powerful computer However early concerns that the computationally more intensive first option might introduce unacceptable time delays proved to be unfounded All prototype development used the onboard GPS option Wireless Communications Transceiver The system uses Cellular Digital Packet Data CDPD wireless transmission technology CDPD is a cutting edge wireless communications protocol which allows direct connection of the remote devices to the Internet Embedded Microprocessor System performance is controlled by an embedded single chip microcomputer S
42. he output from this device will then be fed to the lotech DBK11A screw terminal card This is the device that will receive the conditioned amplified signal and in turn connect with the computer via a parallel port sending the data to DasyLab This data was then imported into Matlab for display Figure 7 10 shows a block diagram of the complete system 42 Results Figure 7 9 Accelerometer mounted to testing plate Data from Accelerometer Crash Test PC Simulator Power Daisy With Lab accompanying Matlab Accelerometer Accelerometer Figure 7 10 Schematic for data acquisition After construction of the system a series of twelve tests were conducted Tests were conducted with and without Teflon pads and with and without the Mobile Unit Enclosure A table documenting all the tests is shown in Table 7 1 An X denotes the use of the part an denotes the lack of the part 43 Table 7 1 List of Mini Sled Tests Performed Trial Teflon ACN Pads Box UJ Y 44 5 3 G s F i 5L iil 1 8 Ta FT P ET i 25 0 05 0 1 0 15 0 2 0 25 Tie 8 Time s a Without Teflon Pads b With Teflon Pads Figure 7 11 Acceleration Pulses without the Mobile Unit Enclosure G s 7 G s 0 05 01 045 02 025 03 035 04 Without Teflon Pads b With Teflon pads Figure 7 12 Acceleration Pulses with the Mobile Unit Enclosure 45 Graphs were
43. he same structural cage which protects the occupants Note that there is some degree of overlap between the Mobile Unit and components in late model cars Since the early 1990 all passenger vehicles sold in the U S have been required to have airbags Increasingly the sensors used in these systems are electronic sensors of the type used in this program However modification or connection to the airbag or any other safety systems of the car has been strictly avoided in the Mobile Unit for liability reasons Eventually automakers may choose to use the airbag sensor to drive an ACN systems of the type described here However the Mobile Unit has been designed to be completely independent of all in vehicle systems with the exception of the car battery GPS Sensor Embedded Crash Sensor Embedded Wireless Microprocessor Modem Power Figure 3 1 System Architecture Base Station The Base Station system will 1 receive the emergency call over the Mobile Unit 2 receive GPS data and the crash pulse from the crash site and 3 display the location and severity of the crash using computerized maps for Emergency Response Team dispatch The prototype Base Station will 1 serve as a test bed for later development into a full featured Base Station in later phases of the project and 2 for checkout of the prototype in vehicle device proposed here Note that this system is intended only for laboratory use it is not intended for use a
44. ill not be included in the production prototype During the test the car with installed Mobile Unit was driven on a 10 mile circuit around Rowan University From the continuously updated map on the Base Station we were able to track the student team as they drove from street to street and were able to even identify which lot they parked in upon their return Low Severity Impact Testing Objectives The ANJEL Mobile Unit was tested in an impact test for two purposes 1 To check if the unit can detect a crash 1 ensure the system can survive crash The goal was to evaluate the performance of the Mobile Unit in low severity crashes For this project low severity was defined as that impact speed at which the airbag would normally deploy approximately 12 15 mph Higher severity crashes such as the NHTSA full barrier 30 mph crash tests are expected to result in peak decelerations of 30G or higher Although the contractual requirements of the current project are limited to evaluation of the Mobile Unit at low severity crashes it is recommended that follow on projects test the unit in higher severity crash tests Micro drop Test A micro drop tower design was chosen as the first impactor because of the simplicity of its design and ease of fabrication The micro drop tower shown in figure 7 1 was constructed by cantilevering a rope and pulley system from the top of the Rowan Drop Tower The Rowan Drop Tower normally used in aircraf
45. ingle chip microcomputers such as the MicroChip series Z World series or Motorola 68HC12 series combine onboard memory reasonable clock rates and onboard A D capability into a low cost package which is readily interfaced to sensors such as those used in the ANJEL system Power Power for this system is provided by the passenger car 12 volt electrical system Note that per our design guidelines this is the only interconnection between Mobile Unit and the passenger car Power from the car battery will be conditioned as necessary before input to the Mobile Unit electronics Storage of backup power small onboard battery permits successful operation of the Mobile Unit even if car battery power is lost as a consequence of the crash Crash Algorithm A crash algorithm a software module in the microprocessor was developed to detect a crash while avoiding false alarms The Mobile Unit must be able to distinguish between actual crashes and low severity crashes or non crashes such as panic braking or backing into a shopping cart To detecta crash the microprocessor samples the accelerometer output at 1000 Hz 1 sample per millisecond Based upon examination of National Highway Traffic Safety Administration crash tests coupled with crash test modeling the crash detection algorithm was designed to signal that a crash has occurred if a 10 miles hour change in velocity occurs in under 50 milliseconds To put these time intervals in perspectiv
46. lTOut 0001 FinTOut 15 Model Information VehID INTREPID Make DODGE Model INTREPID Year 1999 Wt 1749 IniVel 240 23 MasslID OccComp Descr Occupant Compartment Wt 1331 278 MassID Engine Descr Engine Wt 267 7219 MassID Wheels Descr Front Wheels Suspension Wt 150 spriD Occ Bar Descr Occ Bar NegMass OccComp PosMass Barrier StaType SI SU 15410 1 ST 0 X 0 39 47368 78 94737 118 4211 157 8947 197 3684 236 8421 276 3158 31547095 355 2632 394 7368 434 2105 473 6842 5B L579 552 6316 5924 LOSS 631 5789 671 0526 710 5263 750 F 0 9266 408 26253 13 61447 73 95652 23 114329 7 105060 7 89819 73 97692 63 119951 4 128010 6 152992 9 179145 188187 218022 6 317113 6 218785 2 203 WEG aay 188647 Tui T DynType AM 51 02251986 243 0128 SprID Radiator Descr Radiator NegMass Engine PosMass Barrier StaType SI SU 1028 647 ST 0 X 0 32 14286 64 28571 96 42857 128 5714 160 7143 192 8571 225 257 1429 289 2857 321 4286 353 5714 385 7143 417 8571 450 F 0 0 0 0 0 403 2021 1381 185 2854 973 4759 486 7668 291 11635 49 22504 78 22504 78 22504 78 22683 99 51 1 244783 MMax 22 17755 15 Descr Wheels Bar NegMass Wheels PosMass Barrier StaType SI SU 21114 ST 0 X 0 39 28571 78 57143 117 8571 157 1429 196 4286 235 7143 275 314 2857 353 5714 392 8571 432 1429 471 4286 510 7143 550 F 0 4361605 4361605 3 291417 13 88973 36 24468 76 82959 186 8221 186 8221 186 8221 186 8221 249 383
47. mand line runwatch hitwd hold value 5 1 begin conversion to A D for A D to Tinish converting FOr A D input 605413 release held value oOutport csoT4 2 compute hex values upper input 16 lower fmod input 16 Jy assign hex values output 0 reference upper bia 22 rererence lower ls Tg write value to serial port 214 outp t 61 2 CDPD_GPS cpp This program sends data through the CDPD This program takes the serial data from GPS in from port 0 L4 Bao OU Do Lor Ene CDPD define IBAUDO 9600 1200 baud rate define IBAUD1 9600 1200 with modem either 2400 or 1200 without modem gt 19200 9600 4800 etc define TBUFSIZE 384 I4 Size Of tran omit burtfers define RBUFSIZE 384 lf eize Of buffer MODE 24 Lf 8 data no parity l stop char NO MODEM 0 we don t want modem char ECHO 1 we do want character echo main 3 ING char tbut TBUFSIZE buffer ehar 17 receive buffer char buf RBUFSIZE 1 dummy buffer for receiving a complete command char burtz RBURSIABTI z allows serial port 0 to work if ROM reload iDZ70 2b fendif communication with Dynamic C is lost
48. n Server Perl usr bin perl mode COM2 BAUD 19200 PARITY N DATA 8 RETRY N 1 SCOUNTER 0 Slogfile C acn crash log Scrashfile c acn crash txt Set to 2 when an alert happens Snemafile c acn gps nmea Logs all valid NEMA with GPS data che 1077 Schris Chir 9 Takes a TCP connection echos it back but several now user TOs Use TO roe lecti open LOGGER gt gt Slogfile amp log WARNING unable to log status msgs open NMEA gt gt Snemafile amp log WARNING unable to log NEMA String iy listener TO tfSOOCket itINBET c2Hew Protoe top LocalPort gt 91117 Listen gt 10 Timeout gt 500 Normal connect timeout 60 or die not open port 9111 for listening Sselector new IO Select listener J STARTUP Listening pork 9111 TCP Tor APRS systems Set up UDP port Add to selector here 10 Socket INET gt new LocalPort gt 9110 Proto gt Reuse gt 1 or die Can not open UDP socket Sselector gt add Sudp flog STARTUPS Listening to pott 9110 UDP for Al rts whale consready Sselector gt can read 4 for my Sconnection consready if Sconnection Slistener Is it a new TCP connection my Snewconn Sconnection gt accept Say hi to our new friend S5newconn autoflush 1 print newconn f ROWAN ECE PROJECT EXPERMENTAL S
49. n appendix to this report As shown in Figure 4 1 this model uses a system of non linear springs and lump masses to simulate a vehicle s structural response during a collision Using this model a crash simulation was conducted Figure 4 2 shows the outputs from the program in terms of acceleration pulses for the impact These pulses were integrated to find the change in velocity of the vehicle These simulations were then repeated using a range of speeds varying from 25 70 mph By analyzing the results from each simulation the time for a 10 km h change in velocity during a crash was found Figure 4 3 shows that for the simulations run the maximum time required for a 10 km h change in velocity was 36 milliseconds If the driver in a car traveling at 60 mph slams on the brakes it takes about 500 milliseconds to undergo a 10 km h change in velocity Clearly this method successfully differentiates between a crash like situation slamming on the brakes versus actual crash Although this algorithm should be adequate for field testing of the ANJEL system additional test and simulation data should be evaluated prior to development of an algorithm for a production ACN system The final algorithm should set crash no crash threshold based on additional crash configuration including different vehicle makes and models side impacts vehicle to vehicle impacts and vehicle to rigid barrier impacts 5 MOBILE UNIT SYSTEM DESCRIPTION System Architect
50. n detecting a crash Peak deceleration during the test was observed to be 9 Gs Future Work Testing of the a Mini Sled Impactor System showed the potential for improved performance especially in three areas e Rails Modifications If an air cushion were used to carry the system friction could be greatly reduced 46 Carriage Modifications Bearings rollers wheels should be added to facilitate motion of the Mobile Unit Air Cylinder Modifications To improve the speeds achieved more air flow should be allowed to leave the piston on the outstroke 47 8 CONCLUSIONS This project has developed Low Cost Automated Crash Notification System for eventual field testing on New Jersey highways The system was developed in response to national studies which show that nearly half of all traffic crash fatalities occur before the crash victim reaches a trauma center Many of these deaths can be attributed to the inability of EMS personnel to locate and reach the victim during the so called Golden Hour after the accident when emergency medical treatment is most effective The goal of this project was to dramatically reduce EMS response time by developing and testing an advanced in vehicle system that automatically transmits the location and severity of a crash to EMS personnel Specific accomplishments of the project include e The project has designed developed and tested a low cost functional system that combines wireless communi
51. ned from the car cigarette lighter adapter rather than directly connecting to the 5 electrical system The second prototype referred to as Rev B in this document would be designed using a fully functional Rev A prototype as a starting point While maintaining the functionality of Rev A Rev B would explore the possibility of lower cost approaches to Automated Crash Notification The objective was to design build and test a second prototype which could be fabricated in quantities suitable for field testing in New Jersey 4 2 Base Station Development Approach To test both of these prototypes a Base Station was developed which could field calls from the Mobile Units and simulate the operation of a future automated crash notification 9 1 1 center A key objective of the Base Station was to provide a means to test Mobile Unit wireless communication i e to receive ACN messages from the Mobile Units and to plot the location of these Mobile Units on a computerized map A second objective was for the Base Station to serve as the test bed for evaluating automated mapping products To limit development costs the research team sought to use commercial off the shelf software and mapping products whenever possible Note that the Base Station developed under this project was intended solely as a means to test correct operation of the Mobile Unit While it is hoped that our Base Station design may provide some guidance for future 911 cen
52. ng thePacket getData 0 thePacket getLength oystem out println thePacket getAddress at port thePacket getPort says 5 reset the length for the next packet thePacket setLength buffer length catch Exception oystem err println 52 A 2 Simulated Mobile Unit Java Function The following code was developed to allow the Base Station to be tested from a simulated Mobile Unit The simulated Mobile Unit was a Java program running on a separate PC This program transmitted UDP messages containing GPS coordinates to the Base Station Server import java net import java io public class Key2UDP public static void main String args system out printin Mobile Unit Simulatorin String hostname 150 250 105 127 Address of ACN Server int port 8080 Setup Datagram Socket try InetAddress dest InetAddress getByName hostname oystem out println Destination dest DatagramSocket soc new DatagramSocket Read keyboard messages and send them to the UDP Socket BufferedReader in new BufferedReader new InputStreamReader System in while true string msg in readLine if msg equals break byte msgbytes msg getBytes DatagramPacket theOutput new DatagramPacket msgbytes msgbytes length dest port soc send theOutput catch Exception e oystem out println 53 Base Statio
53. nverter to a specific memory location 0 4104 The micro controller then assigns value of 1 to this address This assignment serves as instruction to the ADC to hold the voltage sample that it currently sees on its X axis input Following this a value of O is written to the address 0x40C1 which initiates analog to digital conversion When the conversion is completed the ADC sends an interrupt signal to the microcontroller It should be noted that the value on the data bus will hence represent the acceleration reading at that particular instant of time After the interrupt signal a value of 1 is reassigned to the address 0x4104 This tells the ADC to release the held value of voltage The acceleration reading acquired by the micro controller is then stored in the circular data buffer It is compared with preceding values to ascertain whether a crash has occurred or not If a crash is detected the data buffer is written out through the microcontroller s serial I O channel if no crash has occurred the above loop is repeated The C code for algorithm is shown in the file crastest cpp included in Appendix C at the end of this report The microcontroller continuously logs acceleration values from the accelerometer every millisecond It monitors the data in 40 millisecond pieces to determine if a crash has occurred It also updates the vehicle s location from the GPS unit every second The software implemented for this is primarily inter
54. ototype and a prototype Base Station has been successfully tested in a series of laboratory low severity impact tests Although system performance has shown unusual promise it must be emphasized that the systems tests have been conducted solely in a controlled laboratory setting Prior to development of a production system the following additional tasks are recommended Field Testing Future work should include a second research phase which will perform operational field testing of the ACN system A fleet test would evaluate the performance of the system in both crash and non crash modes and would provide important consumer acceptance feedback from the motorists A fleet of 1000 ACN equipped cars could be expected to incur approximately 10 collisions per year for evaluation of the system under crash conditions The location of the cars should be chosen to produce a fleet mix representative of the New Jersey s mix of urban and rural highways Captive fleets such as those maintained by the New Jersey Department of Transportation or the New Jersey State Fleet would be ideal for such a field test Higher severity impact testing Although the system has been demonstrated to be crashworthy in low severity tests we recommend that follow on projects should conduct additional laboratory performance testing of the system at the higher impact severities attainable in staged crash tests and HyGe sled tests Integration of ACN into existing New Jersey Emerg
55. re as follows 1 8 x 2 05 2180 running at 9 216 MHz O SRAM 1 32 I O 2 Serial Ports 2 DMA Channels 2 Programmable Timers One of the main reasons that this microcontroller was chosen was the support it offered for programming in Dynamic C which is a variant of traditional C This along with the vast library support offered by Z World has helped speed up development time The microprocessor used in Rev B is the PIC17C756A processor It has a clock speed of 33 MHz and is widely used in industry The main features of the chip as listed in the user manual as follows 1 65 x 2 34 68 pin PLCC 4 watchdog timer 12 channel 10 bit ADC 15 preset SRAM 902 bytes 20 One of the advantages of using the micro controller is that is much less expensive than the Z World processor This makes it more suitable for applications such as the ACN where the cost of the final product is important Additionally the PIC17C756A has a 10 bit A D converter which can be used to take in data directly from the accelerometer This eliminates the need for an external ADC and reduces valuable board space requirements leading to a less expensive more compact final product Just as with the Z World processor the PIC chip uses C a widely used programming language Functionality As mentioned earlier the Rev A microcontroller receives input from the A D co
56. ror the SMPO4E which is a sample amp hold chip is used between the accelerometer and the A D unit This will ensure all data points are properly recorded during the detection of a crash Moreover by setting a high sample rate 1000 times per second we can account for the discreteness of the data A D Converter ADC The ACN uses an external voltage comparator to compare readings from the accelerometer to certain threshold voltage values Whenever the voltage exceeds the threshold the ADC chip ADCO809CCNA sends an interrupt signal to the microcontroller which processes this data to determine if a crash has occurred or not The ADC has 8 analog input channels Using the address latch in the ADC the desired input can be converted to an 8 bit digital stream The stream is then assigned to the data bus and read to a particular address location on the micro controller the address used is 0 40 1 This acceleration data is stored in the RAM Random Access Memory in the form of a circular buffer of size 128 If the micro controller detects a crash the contents of the buffer are written out to the RS 232 output from where they can later be acquired by an external device The code for this program is shown in the file DACN cpp which is included in Appendix C at the end of this report Microcontroller T mm aua Figure 5 1 Z World Microcontroller A Z World CM7200 microcontroller was used for Rev A The specifications a
57. rupt driven Both the A D unit and the GPS unit generate interrupts when the microcontroller needs to read a value In the event that the microcontroller detects that a crash has occurred it proceeds to wake up the wireless modem from sleep mode and instructs it to begin emergency transmissions Crash Site Location Subsystem GPS System System Description GPS is one of the only systems available today that can pinpoint one s exact position on the earth anytime in any weather anywhere Twenty four GPS satellites continuously orbit the earth at a height of 11 000 nautical miles These satellites transmit signals that can be detected and used by anyone with a GPS receiver to determine one s location with great precision Consequently the GPS system is an integral part of a crash notification system as it is needed to determine the location of the vehicle during a crash Various companies were researched before a GPS receiver was selected These companies include Motorola www motorola com Trimble www trimble com and SiRF www sirf com Important considerations in the purchase of the GPS receivers include their accuracy locking time for a signal and their ruggedness to vibrations g force and so on All three companies have a host of GPS products that could be used for this application After much consideration the Trimble ACEII GPS core module was chosen for Rev A Since the ACEII is a core module it is designed for OEM application
58. s The specifications of this GPS module are as follows Trimble 1999 Figure 5 2 Trimble ACE II GPS Unit 8 channel continuous tracking receiver Update rate NMEA 1 Hz 25 m 5096 without S A Acquisition typical Cold start 130 seconds 9096 Warm start lt 45 seconds 90 Hot start 20 seconds 9096 Reacquisition after signal loss lt 2 seconds 90 515 m sec maximum Velocity 515 m sec maximum Operating temp 40 C to 80 C Power consumption Primary 5 V DC 5 GPS board only 155 mA 0 78 watts With antenna 180 mA 0 9 watts protocols TSIP binary data NMEA 0183 v2 1 ASCII data TAIP ASCII data The GPS board is capable of outputting coordinates using various I O protocols For our application we will be using the NMEA protocol as this protocol has been standardized One issue of concern here is the power consumption for this unit However since the unit will be running off the car battery this is not a major concern The acquisition times as well as the accuracy of this unit are reasonable Functionality The GPS unit is triggered by the microcontroller as soon as the car is turned on Within minutes the GPS receiver locks onto the satellites and is able to pinpoint the location of the car Thereafter the GPS unit updates the position of the vehicle every second as long as the car is on As result even if the GPS antenna is damaged and the satellite lock is lost during
59. s from being polytropic a key assumption the analysis above Consequently a larger 3 inch bore diameter cylinder was used to ensure adequacy over the test Figure 7 5 shows the air cylinder used for the apparatus 39 Figure 7 5 Air cylinder with 3 bore diameter Despite the conservative design described above initial testing of the apparatus revealed insufficient acceleration of the Mobile Unit To improve the impactor performance the inside of the front plate was bored out to enhance air escape from the non pressurized side of the cylinder This allowed the cylinder to let out the maximum amount of air and completed the construction of a fast acting cylinder Figure 7 6 Test stand with mounted piston 40 Figure 7 7 Rails mounted to test stand Figure 7 8 Carriage on rail system 41 Test Stand and Rails The test stand is one of the most integral parts of the impactor setup It is approximately 11 feet long thereby allowing for 10 foot rails and a 1 foot wide cylinder As shown in Figure 7 7 the test stand holds the cylinder in place and absorbs the recoil from firing During testing the test stand is constrained so that the platform does not slide along the floor Aluminum was chosen as the material of construction In order to ensure that the box experiences the proper acceleration forces it is necessary to guide and support the box with as little interference as possible from the rails Owing to i
60. s a production system Mobile Unit Functional Requirements Crash Detection Crash detection will be performed with an array of accelerometers Detection of frontal impacts requires an accelerometer aligned with the longitudinal axis of the car x axis while detection of side impacts requires an accelerometer aligned with the lateral axis of the car y axis Note that the x axis accelerometer will detect rear impacts in addition to frontal impacts and a single y axis accelerometer will detect both driver and passenger side impacts Angled impacts or frontal offset impacts would detect accelerations along both axes A minimum of two sensors 1 required to detect front side rear angled and offset impacts In the U S these accident modes account for the majority of all accidents The system developed under this program is a two axis system Depending on system cost constraints additional sensors could be added to the system to complement this minimal sensor set Other sensors such as a third sensor in the vertical direction Z axis would provide a complete acceleration time history including vehicle pitching during impact However a review of NHTSA frontal side and frontal offset crash test data suggest that z axis acceleration is negligible compared with the x axis and y axis acceleration It should be noted that the two sensor system cannot detect rollovers Either a dedicated roll sensor or a second sensor in the z axis separated from t
61. s here if ant Slat plat poda sprintf 02d sprinti lat ane islet 00050730 else 21012 olan if int 51 1ong sprinti int Slong sprintf S 05 2f Slong int slong 7 100750590 else Slong sprintf 08 2 Slong 100 Slat e Springer plong Sprint 47500520 my 5 hour mday mon Swday S isdst gmtime time sprintf 7502 502 502 7 Smday Shour my Sstringl 0 TCRIP Szuluslatsns Slongsew LEST EXPERIMENTAL 55 Print Xunsscringlxn s my stringl AL3RT gt APZ100 TCPIP S lat ns Slong ew EXPERIMENTAL CRASH SYS Schrl0schri3 my aprssystems seleotor can write 1 for my Swrcon aprssystems if Slistener 55 5 Sudp 16104 ALERT Notifying APRS at Swrcon gt peerhost print Swrcon ROWAN ECE ACN PROJECT EXPERMENTAL SYSTEM APRS 07 print Swrcon Sstringl End if not TCP or UDP listeners 4 End For my wrcon of aprssystems End if GGA NEMA string else 4 amp log WARNING Wierd UDP string from Sudp gt peerhost Sudp gt peerport amp log WARNING Sudpd
62. s well as side impacts The accelerometer then needs to acquire data in two directions as well e Saturation Car accidents tend to take place over a wide range of impacts For example victims 30 mph accidents may experience up to 30 g s The range of the accelerometer needs to be wide enough to ensure the system does not saturate at low 0 5 For Rev A it was decided to use an Analog Devices ADXL250 dual axis accelerometer The component is small in size measuring a mere 0 4 x 0 3 It possesses the ability to measure accelerations along both the X and Y axis thereby enabling the system to detect both frontal and side impacts Moreover the accelerometer has a range of 50 g Since Rev A is concerned with crashes to about 30 g the system is in no danger of saturating Since Rev has been limited to the detection of frontal crashes readings from the accelerometer are only considered along the X direction The process can be modified however to allow for inclusion of readings along the Y direction Sample and hold Chip The A D converter is continuously connected to the onboard accelerometer However the A D converter samples the accelerometer only at discrete time intervals In addition the analog to digital conversion process requires a finite amount of time Since the voltage inputs to the A D will change continuously failure to hold a voltage sample during the A D process may skew the data To rectify such an er
63. t seat crash testing allows freefall drops from heights up to 6 meters The Mobile Unit enclosure was fixed to a wooden platform which could be hoisted to the 34 desired drop height rope pulley system During free fall the platform and attached Mobile Unit was constrained by four guide cables which passed through four eye hooks attached to the corners of the platform Figure 7 1 Micro drop Tower Apparatus Instrumentation The fixture holding the Mobile Unit during testing was instrumented with an external accelerometer as a check against the internal Mobile Unit accelerometer The major challenge was determining a location for the accelerometer so it would not get damaged A second concern was how the accelerometer on a moving system would be connected to a stationary data acquisition system The accelerometer required a constant power source and a cable for transmitting data back to the A D board The cable was mounted manner to avoid tangling during the drop tests Experimental Setup The accelerometer was connected to a signal conditioning unit through the use of an umbilical cable The signal conditioning unit provided the required 35 excitation voltage to the accelerometer The signal conditioning system also provided signal filtering and amplification prior to input to the data acquisition board To implement the data acquisition system a PC was used with the DasyLab data acquisition software Procedure Th
64. t by the microcontroller as outlined in the following table Table 5 1 Charging Modes _ Output Current A Vset 13 3ISsense Pulse Trickle Vset 13 3Rsense 12 5 Duty Cycle Vser 1 3 3Rsense The only mode that will be used in this design is the pulse trickle mode where Vset 1 145V and Rsense 0 560 so the output current is 154mA with 12 5 duty cycle This mode can be set during start up of the microcontroller 21 Conclusion Figure 5 4 shows a photograph of the completed Rev A prototype Figure 5 5 presents a overall schematic of the system Table 5 1 provides cost estimates for the Rev A system A similar calculation of costs for Rev B estimated Mobile Unit costs in quantities of 1000 at 400 450 per unit GPS CDPD Wlcrocontroller Figure 5 4 Completed ACN Rev A unit 22 Figure 5 5 Basic Schematic of ANJEL Mobile Unit Rev From Antenna 459v 2 98 accu Jk EUR PLUR ch 45 C2 O 1u To Cellular Unit 23 Table 5 2 Cost for ANJEL Mobile Unit Rev ACCEL Bill of Materials design 51 5 Count ComponentName RefDes PatternName Value Description Cost 5 1 CM7200 01 CM7200 Core module 7180 99 00 1 EXPEDITE 2 IDC2 6M Novatel Expedite Modem 230 00 1 ANTENNA CABLE
65. tage levels a voltage regulator is used Each regulator is supported by an EMI electromagnetic interference filter and bypass and bulk capacitors An LM2940 linear voltage regulator provides the 5V DC supply while an LM4040 provides the 5V analog reference An LT1098 sources the 3 6V DC supply The back up battery system is composed of 5AA NiCd batteries and the supporting charge system Each battery has a cell voltage of 1 2V giving the total battery pack a voltage of 6V The 12V input from the car battery and the battery pack outputs are both connected through power diodes to the to the inputs of the three regulators When the 12V from the car battery is removed the battery pack diode becomes forward biased and continues delivering power to the regulators Specifically Schottky diodes are used such that there will be a minimal voltage drop across the diodes p MOS switch installed between the battery pack and its diode is the means for switching battery pack power into and out of the rest of the circuit The p MOS 20 switch is closed when presented with at its gate and open when SV is present at the gate In this way battery pack power is conserved when not needed Under operating conditions there are two circumstances that would result in the loss of the 12V supply to the circuit Either the car has been turned off or the car battery has been disconnected as the result of a crash When the microcontroller senses the loss of
66. ters the current Base Station is in no way intended to serve as a replacement for current 911 dispatch centers 4 3 Crash Determination One of the key functions of an ACN system is its ability to determine whether a crash has occurred or not This makes the design of this sub system very critical One possibility would be to monitor the airbag sensor in the car and trigger the crash notification system in the event that the airbag deploys However modification or connection to the airbag or any other safety system of the car was avoided for liability reasons While automobile manufacturers may eventually choose to use the airbag sensor to drive an ACN system it was decided to monitor the vehicle s acceleration profile to determine whether or not a crash had occurred An Analog Devices ADXL250 dual axis accelerometer was chosen to monitor the acceleration felt by the car This accelerometer was chosen for a number of reasons One main reason was the small size a mere 0 4 x 0 3 Moreover the fact that it is a dual axis accelerometer allows us to monitor the acceleration in both the x and y directions This allows the system to detect both frontal as well as side impacts Also the accelerometer has a range of 50 0 Even in 30 mph accidents the passenger compartment often feels up to 30 05 Hence it is important to make sure that the range of the accelerometer 15 sufficiently broad to avoid saturation During normal operation
67. the 12V supply it checks to see if it is a valid crash state If it is not the microcontroller leaves the p MOS switch open and the battery pack does not supply power to the circuit system shutdown If the microcontroller is in a valid crash state the p MOS switch is closed and the battery pack provides power to the circuit so that there is still power to continue transmitting the crash coordinates The actual control of the p MOS switch is accomplished through the use of a D type flip flop The clock and the D input are directly connected to port D of the microcontroller A state change operation in the D type flip flop requires two occurrences First the data bit at the D input must be set 0 for on 5V for off Second the clock must receive a rising edge On start up the microcontroller sets the output of the D type flip flop to OV closing the p MOS switch and checks to be sure that back up power will be available in the event of 12V supply loss To prevent any leakage current from flowing into the microcontroller 10 resistors are installed between the flip flop inputs and the port D terminals As stated previously the battery pack is supported by a charging network The MAXIM 1640 charging chip is employed to PWM pulse width modulate a 1mH inductor to provide a constant current level for charging the battery pack The charging chip has a two bit interface with the microcontroller at port C This allows the charging mode to be se
68. the EMS center with an early snapshot of the seriousness and potential injury consequences of the accident The message to the Base Station should include both these data facets as well as information detailing the time of the crash and a description of the car Crash location can be as straightforward as the GPS location longitude and latitude Crash severity should be provided for each crash sensor and can be either the delta velocity or the crash pulse along each axis It should be noted that while inclusion of the crash pulse requires transmission of a longer message the crash pulse typically provides sufficient information to infer whether the car struck a tree or another car which may require additional EMS personnel Inclusion of crash severity for each axis allows the Base Station to distinguish between frontal and the potentially more serious side impacts Crash Location The Crash Location is the single most important data facet transmitted by the Mobile Unit In order to extract meaning from the GPS messages sent from the Mobile Unit the form of the data i e binary ASCII delimited continuous etc and the interface it would require modem serial port etc must be clearly defined Current GPS devices provide several different options for GPS coordinate output The most widely used format however is that set by the National Marine Electronics Association NMEA Of the three versions of the NMEA standards that were found
69. the NMEA 0183 was the most recent and workable This standard originally set for marine instrumentation dictates both a 27 data and interface protocol The NMEA transmissions consist of strings of printable ASCII characters carriage returns and line feeds Comma delimited sentences such as the one in Figure 6 1 are sent in succession through the serial port typically at 4800 baud Trimble 1999 Conexant 1999 GPGLL 3744 953 N 12225 319 W 182220 device 37 degrees time ive 44 953 minutes 18 22 20 String af equator A valid 122 degrees Vzinvalid 25 319 minutes West nf nrime meridian indicates NMEA sentence Checksum cumulative XOR sum that the base station should get if sentence Is intact Figure 6 1 An NMEA 0183 sentence A dollar sign indicates the start of each new sentence It is followed by two letters indicating the transmitting device in this case GP indicates a Global Positioning device and then three more letters representing the sentence type Each sentence type has specific fields of known length separated by commas that remain in place even when a field is left empty The GLL sentence shown here carries information about latitude in the second and third comma separated fields and longitude in the fourth and fifth fields Initially the first latitude field looks as if it is divided into two sections at the decimal point when in fact the division occurs after the second di
70. ts simplicity and ease of operation a dry rail system was used For better lubrication the rails were polished to a smooth finish To further reduce friction Teflon pads were mounted on the inside of the carriage brackets Carriage As shown in Figure 7 8 the carriage is supported by two brackets on either side A common base connects the two brackets This platform is approximately 12 inches long and as wide as the rails dictate it to be to reduce the friction between the support brackets and the polished rails felt is glued to the inside A wall is attached to the rear of the carriage so the ACN does not come in direct contact with the piston rod Moreover at the end of the rails there will be a wall so the carriage will not experience a metal on metal collision Instrumentation The data acquisition system consisted of the following components Analog Device 150 single axis accelerometer Figure 7 9 Metraplex Series 300 signal conditioning system lotech DBK11a screw terminal card DasyLab data acquisition software PC Workstation The carriage was instrumented with an ADXL 150 single axis accelerometer The ADXL150 accelerometer was used to measure the acceleration experienced by the Mobile Unit during the impact This accelerometer was connected via an umbilical cable to the Metraplex signal conditioning system The signal conditioning system provides power for the accelerometer and amplifies the return signal T
71. ure The goal for the initial prototype was to investigate the feasibility and workability of the concept involved with the ACN As a result the first prototype also referred to as Rev A adhered to the baselines established for the overall design No major flaws were found during the prototyping efforts and no major changes were necessary Rev A hence manifests the same system architecture and functionality as outlined earlier in Section 3 An advanced prototype Rev B has also been constructed based on the same architecture Rev B is a more advanced slightly less expensive version of the Mobile Unit which corrects minor flaws detected in during Rev A testing Crash Detection Subsystem Silicon Accelerometer The crash algorithm used in the ACN relies on measuring the acceleration of the vehicle see Crash Algorithm Section 4 Consequently the accelerometer becomes a key component of the crash detection subsystem Several factors needed to be considered in determining which accelerometer to use e Size The system should be as compact as possible since it needs to be portable Additionally several physical and spatial constraints are imposed by the potential location of the system in existing cars e Dual axis Although Rev A focuses on the detection of frontal impact the long term goal of the ACN is to be a fully functional crash detection system Consequently the ACN would need to detect two kinds of accidents frontal a
72. us ws nterrupt handler interrupt LoR readAD 4 inport TMDROL status 0x0100 this function calls the CDPD and transmits the gps coordinates void crashtransmission 4 disable all interrupts DI wake up CDPD OULtDorL CcSoTT 1 eontinously transmit while 1 At DU 69 70
73. utputil4 s char 16 fill reference array with some values 63 reference 0 0 rererence 11 tits rererencelz 7272 reference 3 3 4 14 reference 5 5 reference 6 6 7 7 reference 8 8 reference 9 5 7 reference 10 A reference reference 12 C reference 13 D rererence 14 reference 15 runningtotal 0 6375 maxruntotal runningtotal minruntotal runningtotal status 0 coordinate coordinate coordinate OF coordinate coordinate coordinate coordinate J H0 0 0 1 2 3 4 5 6 s Initializes Port 1 of the 7180 for writing Dina Zit bury CDU BUR SIAR MODE IBAUD NO MODEM ECHO 05 index 0 time index 0 crash e for 120 950 FFE circbuffer i 128 runwatch while 1 hitwd Value 64 60554 5202 begin conversion to A D wait for interrupt while status 0x0100 read A D value xDport qoodrl Status 0 0000 release held value oora LU place algorithm here input CLEeCOULrer pr gt Maxruneo clas maxc untobral
74. vanced in vehicle system that determines that a serious automotive collision has occurred and automatically summons Emergency Medical Services EMS response Specifically the proposed project will design develop and test a low cost functional system that combines wireless communications and Global Positioning Systems with a network of inexpensive sensors for crash detection The purpose of the system is not only to shorten the time it takes to notify authorities of the crash event but to improve the quality of the response This project will perform limited field tests of a prototype automated collision notification system ACN A follow on phase of this effort will seek to conduct an operational field test of the ACN system using up to 1000 privately or publicly owned cars in a representative cross section of the State of New Jersey 3 System Requirements Architecture The Automated Crash Notification system developed under this project is referred to as the Automated New Jersey Emergency Locator ANJEL ANJEL is composed of two major subsystems 1 the Mobile Unit which is installed in the occupant compartment of the vehicle and 2 the Base Station which is responsible for receiving distress calls from the Mobile Units and reporting their location to emergency response dispatch personnel This section describes the requirements of each of these subsystems Mobile Unit The Mobile Unit is responsible for detecting a crash determinin
75. when emergency medical treatment is most effective National statistics clearly show that despite a growing wireless communications network and the availability of medivac transport the time to notify emergency personnel of a crash and respond the crash victims can be quite lengthy For fatal crashes in the U S the average pre hospital time is approximately 30 minutes in urban areas and 1 hour in rural areas NHTSA 2000 Currently emergency personnel must rely on passing motorists highway patrols and traffic reporters to report crashes Often the individual reporting the emergency may not know where he or she is let alone be able to direct help to his or her location These delays can be especially lengthy in rural relatively unpopulated areas where a crash site may go undetected for hours and occasionally days Crucial to getting help to a crash victim is prompt notification that a a crash has occurred b the location of the crash and c some measure of the severity or injury causing potential of the collision Automated Crash Notification Systems capable of performing many of these tasks have been installed as expensive options on a limited number of high end luxury cars The OnStar System for example costs 700 for installation carries 200 400 annual fee and is currently offered only for select General Motors models Thomas 2000 The idea behind Automated Crash Notification is to equip cars with a crash sensor whic
76. with existing or expanded 911 systems However this effort will require coordination with existing 911 system operators and careful attention to how best to present crash information graphically to operators who are more accustomed to receiving voice only calls The Base Station developed here will provide an early evaluation of possible 911 operator user interfaces The Base Station may also be suitable for limited field testing of the system for captive fleets such as the State Police or NJDOT vehicles 4 DEVELOPMENT APPROACH The ANJEL system required the development of two major components 1 a Mobile Unit and 2 a Base Station This section describes the development strategy to design build and test each of these components 4 1 Mobile Unit Development Approach The development strategy was to develop the Mobile Unit in two stages The first stage was to demonstrate proof of concept The second stage was to explore designs which would lead to a lower cost Mobile Unit While important reduced cost was to be attempted only after successful proof of concept To attack these two design criteria a series of prototype Mobile Units was planned for development The first prototype Rev A would be designed to demonstrate proof of concept The second prototype Rev B would extend Rev A and would be designed to explore consumer cost reductions Proof of concept required the design fabrication and testing of a prototype Mobile Unit whi

Development of a Low-Cost Automated Crash Notification System

Contents

Download Pdf Manuals

Related Search

Related Contents