TRANSPORT RESEARCH LABORATORY TITLE by The identification, prioritising and analysis of accident blackspots in Mialaysia Radin Umar Radin Sohadi and C J Baguley I Overseas Centre Transport Research Laboratory Crowthome Berkshire United Ki(ngdom 'U ;


13 [2s[P
HIN
Sohadi, R U R and C J Baguley, 1993. The identification, prioritisation and analysis of accident blackspots in Malaysia. In: National Conference on Engineering Education and Research: Achievements and challenges, held at University Pertanian Malaysia. Serdang. Selangor. 15 -16 June 1993. -I THE IDENTIFICATION, PRIORITISING AND ANALYSIS OF ACCIDENT IBLACKSPOTS IN MALAYSIA by RADIN UMAR RADIN SOHADI Dept. of Civil and Environmental Engineering Universiti Pertanian Malaysia and C.J. BAGUELY Overseas Centre Transport Research Laboratory, United Kingdom ABSTRACT This paper presents the accident investigation and treatment framework developed for the Serem ban and Shah Alam Pilot Project. Accident data are based on the revised police accident form P0L27(Pin 1/91) specially designed to enable processing and analysis to be carried out using a customised microcomputer accident analysis package, MAAP. A hazardous location identification and prioritising system based on accident maps, link-node-cell and grid coordinates was developed for urban areas while a kilometre post system was proposed for highways in rural areas. In -depth analyses on selected blackspots are discussed and appropriate countermeasures proposed. 1.0 INTRODUCTION Since 1990 Universiti Pertanian Malaysia (UPM), Ministry of Science, Technology and Environment under the mechanism of IRPA and the National Road Safety Council (MKJR) have been funding research programmes to improve the accident data collection and analysis system in Malaysia. The programmes also aim to encourage wider usage of the system to assist in the identification and effective .treatment. of-accident... blackspots to improve road safety in this country. In view of the massive number of accident records to be analysed (96,500 nationwide in 1991), the use of computer based analysis systems was investigated in early 1990 and a micro computer accident analysis package, MAAP licensed by the Transport Research Laboratory (TRL), United Kingdom, was customised to fulfil the requirement. A pilot project on the diagnosis system for analyses of road accidents was then carried out under the funding of the research sub-committee MKJR with the cooperation of the Royal Malaysian Police (PDRM) and TRL, United Kingdom. The two districts of Shah Alam and Seremban were selected as pilot areas where the P'DRM agreed to cooperate by using a redesigned report form in addition to their current accident form, POL27 (Pin 1987). The new police accident form, POL27 (Pin 90 - Pilot Project), was designed to be easier to complete and incorporated several new items of information of importance to highway engineers; in particular, more accurate location data and collision diagrams. Following appropriate training of relevant police officers in the completion of the new form, it was introduced fully in these two districts from the beginning of 1991. During the course of this trial period several improve- ments to the form were recommended, though these were relatively minor in nature. A revised version of the form known as POL27 (Pin 1/9 1) was then printed in late 1991 for its full-scale implementation in January 1992. An extensive training programme for the police investigation officers over the whole country was carried out prior to 1992. .2&..O MET.HO DOLOGY.- 2.1 Theoretical Framework In this phase the framework of accident investigation and treatment was developed. The accident analysis process involves the identification of accident blackspot locations, establishing general patterns of accidents, analysis of the factors involved, site studies, implementation of countermeasures and evaluation of their effectiveness. The overall process is summarised schematically in Figure 1() The level of 20 RAAJunlJnay19REAAA Journal/ January 1994 / ~~~~~~~~~~~~G,.ndFather I I ~~~MICRO LEVEL I ~~~~~~~~~In-Depth Investigation 1st Level nv. I ~~~~~~~~TraFF;c En9r,. Proagrammer, System Analyst, Supervisor I ~~~~~Suggest Dominant SpperoahSed ik-o I ~~~~~Treatnent Factors SueElvtosCk I I~ ~ ~ ~~~~~~~~~~eileIseto cn~~~~ l~mpkementat;o Figure 1: Blackspot Investigation and Treatment Process REAAA Journal / January 1994 221 investigations and relevant authorities involved are also outlined. 2.2 Computer Program Development Advances in computer technology have been employed to enable efficient interpretation and analysis of large quantities of accident data. The data input, processing and analysis framework adopted is summarised schemnatically in Figure 2(2. Specification files, text files and pointers in MAAP were modified and customised to match exactly with the accident information in POL27 (Pin 1/91). This data were keyed-in directly to the microcomputer using the first MAAP programme option NEWACCS. Geographical maps have been spatially digitised using the Autocad Release 11 software and an Al size digitiser. The digitised data were later used for the graphical display of accident information using either the MAPINFO Geographical Information System (GIS) or the new version of MAAP (Ver. 5.0 Prototype). In the latter case, the digitised data were converted from .DWG format to a simplified .DCM format. Figure 2: Data Input and Processing Framework There are three basic types of standard cross tabula- tions available in MAAP: accident information, injury information and vehicles involved in the accidents. Graphical output can be obtained directly using a special interface with the commercial Quattro Spread- sheet package. If a detailed analysis is required on a particular subject (for example,. accidents., inv.olving,,a,.., motorcycle at dawn or dusk) appropriate conditions can be specified during analysis using the non- standard cross tabulations option. 2.3 Accident Location System A location identification system was developed based on road maps. The link-node-cell and coordinates system (Figure 3) have been proposed and used in the pilot urban areas, while a kilometre post system has been used on the rural highways. To identify the most dangerous spots, the DETAILED ACCIDENT INVESTIGATION program option of MAAP was used to plot and list the accident blackspots according to the required rankings. The list can be based on either the kilometre posts, junctions, links between inter- sections or using the grid coordinates. 3.0 RESULT AND DISCUSSIONS 3.1 Accident Blackspots and Site Prioritising In this pilot project, accidents in Seremban and Shah Adamn have been given their location code by map grid coordinates, node-link-cell numbers and kilometre posts. Ranking of blackspots could now be carried out using any of the above methods. Each method has its own merits depending on the degree of accuracy and/or speed of analysis required. 3.1.1 Accident Maps Figure 4 shows the distribution of accidents in Shah Alamn where the size of circle is proportional to the frequency of accidents. The actual frequency of accidents at each blackspot can be displayed by zooming into the relevant window of the map as each accident is represented by a concentric circle (Figure 5). Inspection of the distribution and size of concentric circles can then be carried out and this gives a rapid visual identification of blackspots in specific areas. Fi.gure..4: -Accident, Blackspots in Shah Alam Using Grid Coordinates 3.1.2 Nodal Analysis Besides accident maps, accidents within 20 metres from major junctions can also be displayed using the nodal analysis options. The above ranking can be cross tabulated using the non-standard cross tabulations option of MAAP with appropriate nodal conditions specified for further ranking analysis as shown in Table 1. An accident point system based on weightings 22 REAAA Journal I January 1994 r--- Hot1 Acc-eo ~ ~~~ -LIZ~~~~~~~' --1 iI_ 22 REAAA journal/ January 1994 3021 3 02 0 7 Z k 3019 3018G,- 3015 30 14 30/2 H~en ja a 3 0 1 1 _ __ _ __ _ 3010 _ __ 4378 4379 4380 4381 4382 4383 438z Figure 3: Location Identification System in Seremban 1 4385 4386 REAAA Journal I January 1994 23 REAAA Journal/ January 1994 23 Table 1: Nodal Analysis in Seremban o t o o ..% fla :S M0 L M 1 R c i c t F l s S l 12 o … ----- R oute non Co ndit ions :non Figure 5: Relative Magnitude of Accident Blackspots within a Selected Area of Shah Alam adopted by the Highway Planning Unit can be used to compute the site priorities where 6.0, 3.0, 0.8 and 0.2 points are assigned to fatal, hospitalised, minor and damage-only accidents respectively. In addition, accident costs based on ESCAP figures (3) could also be used to compute the economic loss due to accidents to give a blackspots ranking for future economic analysis. Similarly, link and cell analyses could also be carried out based on the above methods for stretches between two major junctions (mid-block accidents) or areas. 3.1.3 Kilometre Post Analysis For accidents along the State and Federal roads where kilometre posts are installed, histograms of accidents along a selected route can be plotted as shown in Figure 6. Histograms of accidents within selected stretches which have high accident concentrations can also be produced by further dividing the stretch into smaller sectors, each of 100 meter length. Ranking of accidents within sections of road at 100 meter intervals can also be produced using M1AAP. KRLOME9RES accidets 0 2 -.9 7 aXa....X .1.9 13 casxxsxxxxa.. 2.9 106 0X.xxxxx ...... x5 xxxxx5550050xx..xxx ... x..xx5Oxx-.xx osssasoaaass~xx..xxxxoossosaasasaccasxa.xoccnaaasos 3 -3.9 128 .5a" x 5 0 5 5 0 5 0 0 0 0 5 5 00.. .5 5 5 5 5 0 I ( 0 5 5 0 0 0 5 5 5 5 5 0 0 0 5 0 0 0x 0 5 0 5x 5x5550000000xxx500001t0...5fl5005005055000xx00050000005~x 4 -4 .9 54 000)ttt555500000055500050005000 ~ ~XX X X X XXX50 5 0 0J x x x X5 3 5.9 24 caxxxx ... xxxsxxxsxcxsxx . -6 .9 2 0,cnoccno ... --. nc o c c n o.... .. 7 -7.9 18 xcascaaX a- ass.... 8 -8.9 24 ..... x~x ... xxsxxxxsxxn 9 -9.9 31 00sax sox mxx sxxxsxnoxsxxx..xax... 10 -10.9 20 xxasxsxxxxxxaxxx..xx 1 1 -1 7.9 2 2 s c c s c c c a s s a cx x s s s c s s sx 1 Z 12 .9 73 aaooassssax ssx~ 13 -13 2 xx Tot.ic 492 Fatak x Injury o, Da-os Listing of worst lO0ts 9ectio.. Kitss-ten Post 3.0 Kitotoete Psst 2.7 Klits-st- Post 2.0 Kitonetre Post 3.1 Kit -stre Post 4.8 Kilot tte Post 11.2 Kitoo-tle Post 3.2 Kit etre Post 8.0 KItsn.trn Post 9.6 Kliso-tet Psst 6.4 51so464 123191 71 71614 ..cidents accidentsacc ident S occidents.c scl-detsoccidentsaccitdentsaccidentsacci1dents Figure 6: Histogram of Accidents Along Selected Route in Seremban 3.2 In-Depth Analyses For each blackspot selected, detailed analysis using stick and collision diagrams must be carried out to search for patterns of accident that are likely to be reduced by engineering treatment. Whenever possible, site studies should be organised to collect supple- mentary data such as traffic conflicts, flow, origin- destination, speeds etc in an attempt to obtain further indications of factors leading to road user conflicts which in some events have resulted in accidents. 24 REAAA Journal I January 1994 Node ACCIDENT SEVERITY Accident Junction Name Damages Points Costs No. Accident Fatal Serious Slight 225 Yam Tuan/Sh. Ahmad 2 0 11 55 31.8 RM547,500 214 D.B. TunggaVSh. Ahmad 1 3 8 48 31.0 RM387,000 215 D.B. TunggaVLee F. Yee 0 4 12 27 27.0 RM280,000 313 ZaabalD. Linggi 0 4 9 i 22 23.6 RM228,000 224 Yam Tuan/Jin. Berhala 1 1 3 14 20.2 RM231,000 24 REAAA Journal/ January 1994 Print Dote: Mon 04-05-1993 Univeroilt Perton'lon Moloyslo Stick DIAGRAM ANALYSIS ACCIDENT RECORD FILES: M0W225 COROSTIOND SET: Node 1 Mode 2 * 225* ODD I 2 13 15 A N 18 9 1 101 III 121 131 141 151 161 Ill 101 BIN 109 10 jO9 112 108 108 108 110 110 1'0 101 ill ilI Ill Ill ill 103 103 I IBM 107 119 130 102 101 11 116 I13 120 123 103 107 Ill 115 129 130 104 109 I MRSl 1 71 51 21 2 15 1 7 16 1 111 14 I' '1 1 1 16 16 17 12 1 7I NSA 116 110 j12 120 115 109 115 109 118 109 109 121 I17 101 120 Ill 113 112 I Pi151 1 111111 1 1 111 11 1 11 1 PRlI I D 1 D 1 D j 01 D j 1D 1D D 0 D0 0 0 D D jD 1 OK ..IRK Ink O I DK IK I KINk JOK IN K IRK I RK Bk INK JO IRK IRK IRKlo I NK IRK AwIl 11 1111 I1 111111 of'I1 l1 [o 1 11 1 011s1 S 100 1 1s11 SIs1 1Is1 1 11 11 D-I)D ISSI l 10011 l"I" I " 11 11 1 11 R101 1 11 1111 11 11 1 111 11 1 DOI .. 1.. 1.1.. 1.... 1.... 1.... 1 1... 1.... 1 1.1.1.. . I 191 201 211 221 231 241 251 261 271 281 291 301 311 321 331 34I 351 361 BIN 103 104 104 104 104 106 106 105 105 105 107 107 101 101 102 102 102 102 1 fRll 127 105 lOS 112 130 103 109 103 lID 130 103 Ill l"5 114 103 106 110 112 I HBil 16 16 1613 12 11 16 16 1 5 14 151 1 131 2 1$ 1 214 1 NSA 1'5 120 110 121 I17 112 116 113 116 113 120 120 113 116 101 118 11$ 119 I PEij I IPEJI III 1 PJ 1 11 11 11 11 1 PENA 0 I 0 I I 0 0 0 0 5 0 1 10 0 5 I 0 5 0 1 OK IRK InK jilk IDK IRK ORK INK IOK IRK lRK INK INK INK lRK 10k IRK IRK IRK I AN]- 1*I 1 11 11 1111 11 11 11 1 ssI I IS I IS I Is I I Iss Is IS I I=l I I>=II 001 111 11 1 1001111 1111O1 1 ORKI 11 1 11 1111111 1 11 1 BLM. BULAR TRH. TARSKN NRI= OARS MSA= MASA PEa- PEJ.AKS PRIN KEPARANAN BK = BASK ,¼- EAKAOM Y>- REAR END =.I= SIDE AMI. ANIMAL 01 * OVERTURN SS * SIDESWIPE D&D. SUDUNSENJA D0K' DARK =- .==.==.... I 371 381 391 401 411 421 431 44I 45I 461 "I1 48I 491 501 511 $21 531 5"I BIN 102 103 10 103 103 10 104 104 104 10 10 10 106 106 107 lOB ION I1'8 I IRA 129 jIO 114 115 119 126 101 106 120 102 108 114 127 129 124 lOS 123 130 I ABS 1 71317 111 5 151 4121 2171 61 11214 1 61' 411111 MON 112 108 100 115 121 118 117 116 113 110 107 116 110 117 112 [ill Ill 114 I PE111 1111111111IPJ11 1>'1I Aw1o11 11111111111111 11 IssI ]s1 I ISSID I ItDISt ISISDI s ISS IOSIII DRK1011 1111111111111 111 I "51 561 571 581 59 601 611 621 631 641 651 661 671 681 691 701 711 721 DIR 109 109 110 110 ISO lID 110 il Ill Ill 11 Ill 112 112 I I I I I ION 106 106 J06 DO8 11 113 123 115 116 109 122 114 122 Ill I I I A51 1111 3 141 31 31 6j1 121 211171 31 6 11111 NSA 123 113 113 114 1ll ill 114 I08 106 Ill 114 120 117 122 I I I PEJl 11 1111 IPEal 1111 IPEl 11 11 OK IR IR IR IR IR IR I K IR IR IR IR IR IR IR IK I I I ANSI 1111 11111O 111111 111 DR1011 111 1 Iso 11111 1 111 …. . . . . . . . . . . . . . . . . . . --…=== ===~~~~~~~..=c. .. ..... Stick Mutter * AccIdent 'Code Mutter. .- ' -111--1. - ..- -1-..... - -. 11 1 '021139 2 .022084 3 =023430 4 =030387 5 .019100 6 =019730 7 =020055 8 =025006 41=009150 42=009674 43=010265 44=010452 45.011281 46-012372 47=012970 48'0565331 9 .025632 10.025898 11.000253 12-027504 13'027706 14-028153 15=029898 16z030053 49=017877 50-018113 51=021177 52=023359 53=026505 54=027712 55=028721 56=02879? 17.005640 18.006068 19%.007627 20=008423 21=008420 22=009096 23=01027 24=012604 57.033104 58=033357 59=034229 60=034232 61=035419 62=037828 63=037829 64c037219 25-013342 26=010386 27=010793 28=012139 29=015955 30.016991 31.000659 32=001813 65=038520 66=037699 67.041581 68=040492 33.004612 34.005113 35=005466 36=005626 37-007209 30-000200 39=000617 40-008743 Figure 7: Stick Diagram at Node 225 Seremban REAAA Journal/I January 1994 225 3.2.1 Example of an In-Depth Analysis at Node 225 Yam Tuan/Sheikh Ahmad Junction Using the available database of 1991 and 1992, it can be seen from Table 1 that the worst blackspot in the Seremban pilot area is node 225. This site is a T-junction on the North-South Federal Road F0001 passing through the middle of Seremban town. The majority of accidents at this site occurred on the major one-way road as shown in the stick and collision diagrams in Figures 7 and 8. Side swipe accidents Figure 8: Collision Diagram at Node 225 Seremban involving vehicles merging from the side road, Jalan Sheikh Ahmad, constitute the majority of collisions and a total of three injury and one fatal accidents were reported. Another major type of accident involving injury arises from vehicle-pedestrian collisions (one fatal and four injuries) and, at present, there are no special facilities for pedestrians to cross this busy one- way street. A one-day site study was carried out to capture the near misses, approach speeds, vehicle and pedestrian flows and their manoeuvres. The vehicle and pedestrian flow counts are shown in Figure 9 and the conflicts recorded over the six-hour period are summarised in Table 2. It can be seen that there are relatively large numbers of pedestrians crossing the busy Jalan Yam Tuan (maximum of 890 peds/h) and yet no crossing facilities have been provided. It can be seen also that the merge type of conflict is most prevalent immediately after the Jalan Sheikh Ahmad junction. This is due to the high merging flow which tends not to give way to vehicles from the major stream, often crossing to the left-hand side to Jalan Khalsa as soon as they have joined the main road. Pedestrian conflicts are relatively frequent especially when business activity begins. Approach speeds along the major road were found not to be excessive with a mean speed of 30 km/h and 85th percentile speed of 37 Km/h respectively (Figure 10). 3.3 Proposed Countermeasure In view of high conflicts amongst the merging and through vehicles, strategy to minimise the number of opportunities for conflicts by segregation would be the best option. A raised pavement extension is proposed to begin the segregated lane in the position as shown in Figure 11. A solid delineator is required down- stream of Jalan Sheikh Ahmad so that vehicles joining the main stream will have to form a single file. The channelisation should be extended beyond Jalan Khalsa so that the abrupt crossing conflicts observed will not occur. --Table..2:. -Traffic -Conflicts.-at-.No~de 2,25, -Seremban Conflict Number of Slight and { Serious } conflicts at Junction of Manoeuvre J. Yam Tuan/J. Sheikh Ahmad In hour beginning: 6-hour Type 09:00 10:00 12:00 13:00 16:00 17:00 Total Merging 34 {1 } 29 {2} 38 {3} 39 { 1} 37 38 {1} 215 {8 } Pedestrian 0 1 6 { 1} 8 {1} 2 5 22 {2} Rear-end 6 2 7 8 4 6 33 .Crossing 4 5 3 1 1 2 16 26 ~~~~~~~~~~~~~~~REAAA Journal/ January 1994 26 Hour Car M/C Plcku Lorr Bus 12- 13fPf1 f23 0 3~ Hour Car Mle Pickupla rBus 13 1 1 21 I 309- 1 4517 329) G9 74 60 I8- 1 7163l 6 1 14 j3jJ 11 10-1 l490 379y69 53 53 Total 4 10 155 47 6 5 13-14 5291 27 53 8 0 6-hour all vehs. total = 625 18-1 49 17-I 373 2 6 4 2 5 Jalan Um ~ Total 2701 1987 335 337 343 I-our Ca ~ ik or u 0 7-16 4518 196 5 31 25 1 Total 233 1144 309 151 78 6-hour all vehs. total = 4015 Hour M/c Others Hour M/c Others 1300 [ ~ 1300 [8 90 1600 1 31 5 1700 (j J Total 90 116 Total 173 235 All vehs.= 206 All veh.. 406f 8 ~ Jal n Khals al I l et o Ialan Sheikh Ahmrad Hour M/c Others Hour Mic Others 1300 a 6 1300 12 22 Hour Ca ~ ik or u 1600 [ f ]57 1600 141 18 W9id 1971 4601 sa 8 24 24 76] 1700 [ jj 7. 1700 F 1T 2 0-1 ~ ~1 2541 75 0T~ Total 75 186 Total 36 67 2 -1 4a21 4 1 veh -~2 263 All vehs. - 103 ~7-I 437 23 2 1 4 5 Total 2690 1726 396 156 366 6-hour all vehs. total = 5334 PEDESTRIANS CROSSING IN ARtEA: - Hour A B c 0 09- Id3 3 1[ 47 12-I 1 216I 259 13-I 4 9 j 3291 9 17-1 ] 1 9 Total 899 1186 533 120 6-houal peds toal 373 Figure 9: Pedestrian and Flaw Counts at Node 225 Seremban M .2 --Frequency Speed (Itmlfl)-Cumulative Frequency Figure 10: Distribution of Approach Speed at Node-225 Seremban In order to reduce pedestrian-vehicle conflicts, a pedestrian crossing facilities is strongly recommended. A. pedestrian, bridge-with -appropriate--- guard railing is required to provide a safe crossing facility and to channelise pedestrians to the bridge. Alternatively, if the construction cost of the bridge is too high, a pedestrian refuge may be incorporated in this channelisation scheme as shown in Figure 12. Large road studs along the boundary line of the hatch area and direction arrows on the refuge are suggested to reduce the likelihood of collisions with the new refuge. This arrangement will enable more space to be available for motorcycle parking in the side roads and is, in fact, likely to be more popular with pedestrians than a bridge. Figure 11: Proposed Channelisation with Pedestrian Bridge at Node 225 Seremban 3.4 National Accident Database Following the success of the pilot project in imple- menting a flexible diagnosis system and the launching of the new accident form POL27 (Pin 1/91) in January 1992, it was decided in early 1992 to extend the system to the national level. A downloadin~ transcription programme was written in mid 1992('' and tested with the national accident data available on mainframe computer at Police Headquarters, Bukit Aman. This mainframe computer is used to store all REAAA Journal/I January 1994 2 Nothern Approach--- - 34 6 22 2 A 2 3 4 2 3 2. 2 - 27 Figure 12: Proposed Channelisation and Pedestrian Refuge Option at Node 225 Seremban national data and was designed to produce standard accident tables for the national statistics. Since queries on accident statistics from local authorities, JKR, MKJR etc are frequently non-standard and often very specific, this has now necessitated the writing of special Cobol programs by police personnel to meet these requests. The police are therefore very keen to use the diagnosis system, MAAP, to enable them to respond much quicker to such non-standard queries without the need for special software. A '386' microcomputer has been purchased under this project primarily to perform the task of downloading and converting data formats and to carry out analyses of national accident statistics at Bukit Arnan. An on-line computer linkage using a networking system is currently in the process of being introduced. Accident data will soon be accessible to all agencies responsible for road safety and should thus help these agencies tackle accident problems in a more objective and scientific way. 4.0 CONCLUSIONS An accident diagnosis system and blackspots prioritising system have been developed and tested for diagnosing accident, p~rob.1ems..in.-the pilot.-project- areas. An accident treatment framework has been demonstrated which provides a working model for full implementation throughout Malaysia. The new accident form, POL27 (Pin 1/91) has been redesigned and is now in use nationwide (since January 1992). The computer analysis system developed enables a thorough analysis not only on a macro scale but also on specific blackspots. A hazardous location identification system based on accident maps, link-node-cell system and coordinates has been recommended in the urban areas, with a kilometre post system for the rural highways. In-depth analyses of blackspots can be produced by means of stick and collision diagrams. Appropriate field investigations for capturing supple- mentary data in order to obtain further indications of events leading to accidents have been demonstrated. Diagnosis of actual causes of accidents at a problem location can now be done expeditiously, more easily and more accurately so that appropriate remedial action can be taken. This process also provides the essential "before" data for carrying out qualitative evaluations of the countermeasures implemented. 5.0 ACKNOWLEDGEMENTS The authors would like to thank Universiti Pertanian Malaysia, Ministry of Science, Technology and Environment and the National Road Safety Council for funding the research projects and to Dr. Brian Hills of the Transport Research Laboratory, United Kingdom for his technical support which helped make the project possible. The authors are also grateful to the Royal Malaysian Police; in particular, ACP Supian Ahmad, ACP Abdul Halim, ACP Sycd Abdul Rahman, ASP Acryl Sani and Insp. Roslan, and also to Dr. Amninuddin Adnan of Pengurusan Lebuhraya Berhad and to all research assistants for their support through- out the duration of the project. 6.0 REFERENCES 1. Radin Urnar R.S and Amninuddin A, Accident Investigation Model for Seremban-Port Dickson Highway, Proc. Int. Seminar on Road Safety, JKR/ICA, Kuala Lumpur 1990. 2. Radin Umar R.S, Ahmad Rodzi M. and Aminuddin A, Model Diagnosis Kemnalangan Jalan Raya di Malaysia, Paper accepted to be published in the Journal Sains dan Teknologi Pertanika, 802/92, Universiti Pertanian Malaysia, 1993. 3. Improving Traffic Accident Recording and .Analysis System.In Malaysia, United Nations- Economic and Social Commission foe Asia and the Pacific Report, Thailand 1985. 4. Baguley C.J, TRL Support for MAAP Pilot Project in Malaysia, Technical Paper OU/TP/188/ 92, MKJR/TR1-, Crowthorne, United Kingdom, 1992. 28 REAAA Journal/January 1994 28 REAAA Journal/ January 1994