Moped Rider Violation Behaviors and Moped Safety in · PDF file2 Moped Rider Violation...

Wang, Xu, and Tremont ‐1‐

No: 12-2935 1

Moped Rider Violation Behaviors and Moped Safety in China 2

3 Xuesong Wang*, Ph.D 4

Associate Professor 5 School of Transportation Engineering 6

Tongji University 7 4800 Cao’an Road, 201804, Jiading District, Shanghai, China 8

Phone: +86-21-69583946 9 Fax: +86-21-65982897 10

Email: [email protected] 11 12

Yilun Xu, Research Assistant 13 School of Transportation Engineering 14

Tongji University 15 4800 Cao’an Road, 201804, Jiading District, Shanghai, China 16

Phone: +86-21-69583946 17 Fax: +86-21-65982897 18


Paul J. Tremont, Ph.D 21 Visiting Professor 22 Tongji University 23

4800 Cao’an Road, 201804, Jiading District, Shanghai, China 24 Phone: +86-21-69583946 25 Fax: +86-21-65982897 26


Word count: 4969 (Text) + 6 Figures + 5 Tables = 7719 words 29 *Corresponding author 30

31 32

November 2011 33 34

Prepared for Presentation at the TRB Annual Meeting and Possible Publication in the 35 Journal of the Transportation Research Board 36

37

TRB 2012 Annual Meeting Paper revised from original submittal.


ABSTRACT 1 The purpose of this study is to identify the Moped rider violation behaviors leading to 2 conflicts and crashes and to help guide future countermeasure development. Mopeds (electric 3 bicycles and light motorcycles) are a commonly used personal transportation mode in China, 4 and Moped crashes are increasing rapidly. This increase in crashes may be attributable to 5 certain traffic behaviors of Moped riders. We used video recordings to acquire data on Moped 6 riders’ violation behaviors and the violation behaviors of Bicycle, Tricycle (BTs) and 7 Motorcycle (MCs) riders at signalized intersections. A total of 125 minutes of video was 8 recorded from 10 intersections and rider behaviors were coded from 1455 vehicles. The video 9 data showed that Moped riders commit more violations than riders of bicycles, tricycles, and 10 electric bicycles, and that Moped riders who engage in violation behaviors at intersections are 11 involved in more frequent and more severe conflicts than Moped riders who do not engage in 12 violation behaviors. Typical Moped violation behaviors include running red lights, riding in 13 improper directions, waiting at improper positions, riding in improper lanes, and overloading. 14 Violation behaviors were closely associated with the traffic environment (traffic facilities, 15 traffic flow, traffic signal status and other riders’ behaviors), but not with either the vehicle’s 16 characteristics or the rider’s characteristics. This suggests that countermeasures related to the 17 traffic environment would be more effective than those related to either vehicle or rider 18 characteristics. Countermeasures within the areas of traffic regulation, traffic management, 19 traffic facilities, vehicle management, and rider education are proposed. 20 21 Keywords: Traffic safety, Moped, Violation behaviors, Conflicts, Traffic behavior, 22 Countermeasure 23



1

1 INTRODUCTION 2

Mopeds are inexpensive, two-wheeled vehicles powered either by battery or internal 3 combustion engines under 50cc displacement by the OECD definition. Both electric bicycles 4 and light motorcycles are classified as Mopeds (1), the former being similar in size and 5 weight to bicycles, while the latter are closer to motorcycles in size and weight. Mopeds are 6 able to travel at higher speeds than bicycles, and this may lead to more aggressive riding by 7 Moped riders compared to bicycle riders (2). Mopeds fulfill the trip needs for those who want 8 a higher level of personal mobility than is provided by bicycles, but who are not able to 9 afford private cars (3). The categories of vehicles we are considering in this study are shown 10 in Figure 1 below. This report will adopt the following naming abbreviations when referring 11 to these vehicles. Bicycles and tricycles will be treated as a single category and abbreviated 12 as BTs. Electric bicycles, as well as, light motorcycles, whether electric or gas powered, will 13 be referred to as Mopeds. Motorcycles will be abbreviated as MCs. 14

15 Bicycle Tricycle Electric bicycle 16 (BTs) (BTs) (Mopeds) 17

18

Electric light motorcycle Gasoline light motorcycle Gasoline motorcycle 19 (Mopeds) (Mopeds) (MCs) 20

Figure 1 Vehicle Categories as Defined for this Study 21 22

The Traffic Safety Law of the People’s Republic of China (4), enacted on May 1, 23 2004, classifies electric bicycles as non-motorized, and light motorcycles (both electric and 24 gas powered) as motorized vehicles. This classification based on motorization is important 25 because the law states that electric bicycles (one legally defined category of Moped) must 26 share the non-motorized vehicle lanes with bicycles, and that electric and gasoline light 27 motorcycles (another legally defined category of Moped), as well as motorcycles (defined as 28 two-wheeled vehicles with engines larger than 50cc), must share motorized vehicle lanes. 29 Following the prohibition of most motorcycles in the major cities of China in 30



2004(5), Mopeds have been gradually replacing both bicycles and motorcycles, recently have 1 become the most common personal motorized travel mode in China. (6). In Shanghai, for 2 example, China’s most populous city at 22 million as of 2011, Mopeds comprise over 50% of 3 all riding vehicles (6). This increasing use of Mopeds in China has been accompanied by a 4 sharp increase in Moped crashes. Considering Shanghai alone, 212 people died in Moped 5 crashes in 2010. This amounts to about 20% of all traffic fatalities in the city (7), which 6 would never be neglected. 7 Previous research has identified some rider, vehicle, and operational characteristics 8 of Mopeds that are associated with Moped crashes, but has not focused on violation 9 behaviors of riders as a potential contributor to Moped crashes. Moped rider behavior in this 10 study will usually be contrasted with BT rider behavior. MCs for the most part will not be 11 considered as they represent only 5.3% of the vehicles sample collected in Shanghai. 12 However, there are many more MCs studies than Moped studies available, and while the 13 characteristics of these vehicles are similar, some findings from MCs studies may be 14 generalized to Mopeds. Therefore, some MCs studies have been reviewed, even though MCs 15 are a small part of the two-wheel vehicle population in China. An earlier study, based on 16 1,231 Moped crashes in Mangalore City, India during 2000 to 2004, revealed that the crash 17 rate of men was four times higher than that of women (8). In another study of Moped and 18 motorcycle crashes conducted in France between 1996 and 2005, researchers found that, 19 (besides gender), vehicle type, rider age, helmet use, drunk driving, driving technique, and 20 overloading were important factors associated with Moped crashes (9). It has also been found 21 that motorcyclists or Mopeds drivers who had been drinking were far more likely to be 22 inattentive to the driving task just before they crashed, and to be the primary or sole cause of 23 the crash. Drinking riders were also more likely to be hospitalized and far more likely to be 24 killed than non-drinking riders (10). In another study of crash causes, the hand position of 25 motorcyclists or Mopeds drivers was found to be strongly related to crashes (11). Other 26 human factors, such as driving experience, accident history, education or income have also 27 been found to be significantly associated with motorcycle and Moped safety (12). A survey 28 study of 800 riders in Chongqing, China (13) acquired data on motorcycle and Mopeds 29 riders’ riding behaviors, personality, and traffic safety attitudes, and used a relational model 30 to link these variables to crashes. They found that motorcycle riders’ personality, traffic safety 31 attitude and riding behaviors were all significantly related to crashes. Based on the 32 developing situation of Mopeds in China, it is likely that various Moped violation behaviors, 33 the traffic mix, and unsound and ambiguous Moped traffic regulations are all affecting 34 Moped traffic safety (14). 35

The current study focused on Moped violation behaviors and the conflicts they lead 36 to. The violation behaviors observed in this study were classified according to their temporal 37 and spatial differences and then each violation was coded into one of five categories. Binary 38 Logit models were used to identify and compare the significant factors associated with the 39 various violation behaviors observed. Typical conflict patterns were also determined from the 40 videos. The relationship between violations and conflicts has been found in other settings 41



(15), however, it is an assumption of this study that a positive relationship exists between 1 conflicts and crashes in this setting. This research also identified countermeasures that could 2 be developed to improve the safety of Moped riders. 3

4

2 METHODS 5

2.1 Video Data Collection 6 Moped rider behaviors were videotaped at 12 busy signalized intersections with a high 7 proportion of Moped riders. The video cameras were installed at the intersections from an 8 inclined rear direction, and were capable of recording the operation of all riding vehicles 9 entering and leaving the intersection for about 10 continuous minutes. Table 1 below shows 10 the details of the video data recording schedule, and the specific characteristics of each 11 intersection. Video recording was conducted on weekdays and occurred one or two hours 12 before or after the morning and evening rush hours. A total of 125 minutes of video were 13 recorded at 12 intersections. SONY HDR-CX180E HD Video Cameras were used. These 14 high definition (1440 x1080 resolution) cameras allowed the coder to determine vehicle type, 15 rider characteristics (including gender and approximate age), road facilities and rider 16 behavior characteristics. 17 18

Table 1 Intersection Characteristics and Observation Time for 12 Intersections 19

ID Intersection

Shape

Stop

Line

Separator of

Motor and

Non-motor

Separator of

Non-motor

and Pedestrian

Widened Area

of Waiting

Zone

# of

Crossing

Lanes

Observation

Time

Duration

(sec)

1 Three-leg Y Line Curb Right turn 4 2011/3/4 7:40 720

2 Three-leg Y Curb Barrier None 6 2011/3/4 9:40 775

3 Three-leg Y Barrier Barrier Right turn 4 2011/3/4 10:20 600

4 Four-leg Y Curb Barrier Separated

Waiting Zone

6 2011/3/4 8:15 480

5 Four-leg Y Line Curb None 2 2011/3/4 9:00 560

6 Four-leg N Barrier Line None 4 2011/3/8 11:15 640

7 Four-leg Y None Curb None 2 2011/3/8 11:50 660

8 Four-leg Y Line Curb None 2 2011/3/9 15:20 640

9 Four-leg Y Curb Barrier Right turn 5 2011/3/9 16:00 630

10 Four-leg Y Line Barrier None 1 2011/3/9 16:45 546

11 Four-leg Y Line Barrier None 4 2011/3/9 17:55 498

12 Four-leg Y Curb Barrier None 3 2011/3/10 10:20 757

Note: ‘Y’ stands for Stop Line present, and ‘N’ stands for no Stop Line present; ‘Barrier ’ stands for 3 20 or 4 feet high separator; ‘Separated Waiting Zone’ stands for exclusive non-motorized waiting zone 21 separated from the non-motorized lane. 22 23



The physical layout of the intersection approach, typical of the 12 intersections 1 examined in this study, is shown in Figure 2. Detailed traffic facility and traffic flow features, 2 such as separators for different vehicle types, presence of absence of a stop line, whether the 3 waiting zone was widened, and the number of crossing lanes, were recorded and coded. 4

5

Figure 2 Layout of an Intersection Approach 6

2.2 Video Data Coding 7 Traffic environment, rider characteristics, rider behavior, and conflicts are the four major 8 categories of variables coded. Specific coding details of these four categories are listed in 9 Table 2 below. Please note that rider age was coded into three categories (teenage, middle age, 10 older age), based on the following (subjective) criteria: riders appearing young-looking and 11 with agile motions were coded as teenagers, riders who looked mature, with moderately 12 agile-motions were coded as middle aged, and riders who looked aged and exhibited slower 13



motions were coded as older aged. Violation Behaviors (VBs) were coded into one of five 1 categories. A conflict was coded after a speed or direction change of either or both vehicles 2 was observed when their close approach occurred. Conflict severity (slightly interrupted, 3 moderately interrupted, full stop, or sudden direction change) was also coded according to the 4 nature of the riders’ speed and direction change. 5 6

Table 2 Specific Items in Video Data Coding 7

Category Items Specific Items Explanation Effective width of

non-motorized vehicle lane - Road

facilities Total Moped crossing distance -

Approach split green ratio The ratio of green time and cycle

length at the approach Signal cycle length -

Approach volume of all riding vehicles

-

Volume of motorized vehicles non-motorized vehicles and pedestrians on the crosswise

direction

-

Traffic characteristics

Crosswise traffic saturation time ratio

Total time recorded divided by saturation time of crosswise traffic

Traffic environment

Traffic management

Presence of traffic policemen Yes or no

Age Teenage, middle age, older age Gender Male or female

Rider characteristics

Helmet use Yes or no

Rider operation Moving or stopped and travel

direction

Violation type

Red Light Running, Improper Travel Direction, Improper

Waiting, Improper Lane Use, and Overloading

Rider and behaviors

Rider behaviors

Violation location Lane position and position in the

waiting zone Conflict event Conflict event Yes or no

Conflict Conflict severity

Conflict severity Lightly interrupted, moderately interrupted, full stop, or sudden

direction change

8 Moped conflicts (defined as a change of speed or direction of either vehicle involved 9

in a close approach) could be a negative consequence of violation behaviors (VBs). Based on 10 the video recorded at the 12 intersections in Shanghai, Moped conflicts at intersections were 11 classified as follows: 12

1) Mopeds conflicting with other right turning vehicles; 13 2) Mopeds travelling straight or turning left conflicting with vehicles on the 14

crossing lanes; 15



3) Mopeds turning right or travelling straight conflicting with other through or left 1 turning vehicles; 2

4) Mopeds conflicting head on with other vehicles 3 5) Mopeds conflicting with opposite left turning vehicles; 4 6) Mopeds conflicting with pedestrians or non-motorized vehicles on the 5

crosswalk. 6 These six types of conflicts are shown in Figure 3 below: 7

8

9 Figure 3 Types of Moped Conflicts 10

11 3 RESULTS 12

3.1 Violation Percentages and Behavior Differences between Moped and BT Riders 13 We define the Violation percentage (VP) as the number of all riders observed performing 14 violations divided by the total number of observed riders. This yields an overall VP of 0.511 15



for all vehicles (shown in Figure 1). For Moped riders, the VP was 0.549, and for BTs the VP 1 was 0.459. Apart from overload violations, the VP of Moped ridersdoes not differ 2 significantly from BT riders, even though they differ in certain of the specific violations they 3 commit. However, Violation Behaviors (VBs) by Moped riders are more of a problem than 4 VBs by BT riders because Mopeds currently consist of about 60% of all riding vehicles in 5 Shanghai, and they are increasing faster than BTs. Motorcyclists’ violation behaviors are not 6 compared with Moped riders for two reasons: 1) a small sample of motorcyclists were 7 recorded in this study (only 5.3% of all riding vehicles were motorcycles); 2) motorcycles 8 in Shanghai share roads with cars and trucks rather than with Mopeds or BTs, and this results 9 in little interaction between them. 10 We observed that both Moped and BTs sometimes commit more than one of the five 11 possible violation types on any given occasion. Moped riders are responsible for 71.7% of 12 riding vehicles making two, and 75.0% of riders making three violations at an intersection. 13 This suggests that the characteristics of Moped violations are complex, and that Moped riders 14 differ from BT riders in their violation behavior patterns. VBs for Moped and BTs will be 15 described in detail below. Improper Travel Direction VBs are not covered here, because they 16 are not so common in Shanghai. 17 18 Red Light Running 19 Due to the better power and performance of Mopeds, they are different from BTs s in the 20 characteristics of Red Light Running and Overloading. Moped riders tend to violate red lights 21 further into the red cycle when compared to BT riders. Mopeds are 62.8% of riding vehicles 22 that violate late in the red light phase while they are 48.1% of those that violate early in the 23 red light phase. This finding indicates that signals turning from green to red restrict Mopeds 24 more than BTs, while signals turning from red to green result in Moped riders starting up 25 early compared to BTs. It is because mopeds could start and brake sharper than BTs, which 26 make their driver behave more aggressive. 27 28 Overloading 29 Mopeds, by law in China, are limited to carrying one passenger and a maximum cargo size of 30 (< 1.5 meters high, <0.15 meters wide, <0.2 meters beyond the vehicle’s length). Overloading 31 violations are significantly greater for Moped riders than for BT riders, with Mopeds 32 accounting for 84.8% of all riding vehicle overloading violations. Overloading with teenagers 33 and middle-age persons is the main type of Moped overloading (92%). This may be 34 attributable to the better power and performance of Mopeds compared to BTs. 35 36 Improper Waiting 37 Figure 4 (a) shows that both Mopeds and BT riders tend to wait on the crosswalk for the 38 green lights (beyond the Stop Line). Although proportionally more Mopeds wait beyond the 39 stop line, the data also show there is little difference in improper waiting position between 40 Mopeds and BTs as they both usually wait for the green light together. Figure 4 (b) shows 41



improper lane use positions. Both figures are presented below. 1

2 (a) Improper Waiting Position 3

4 (b) Improper Lane Use Position 5

Figure 4 Positions of Improper Waiting and Improper Lane Use: (a) Improper Waiting 6 Position; (b) Improper Lane Use Position 7

Improper Lane Use 8 The video data suggests that Moped riders regard their vehicles as non-motorized. Figure 4 (b) 9 shows that riding out of the painted edge of the non-motorized lane (on motorized lanes) is 10 the most common Moped improper lane use. As stated above, the Traffic Safety Law of the 11



People’s Republic of China (4) classifies light motorcycles as motorized vehicles and electric 1 bicycles as non-motorized vehicles. But under this classification, fully 80.5% of light 2 motorcyclists are actually illegally sharing the non-motorized vehicle lane with BTs. It is 3 possible most light motorcycle riders in Shanghai do not view their Mopeds as motorized 4 vehicles subject to the specific legal requirements of motorized vehicles (i.e., requirement for 5 a license tag, rider license, and helmet use). This could explain the extensive use of improper 6 lanes by Mopeds found in this study, along with the reason that it would be safer for mopeds 7 riding in non-motorized lane than in motorized lane. 8 9 Helmet Use 10 Apart from these typical VBs acquired from video, helmet use percentage was also recorded 11 because of its importance for rider safety. The Traffic Safety Law of the People’s Republic of 12 China (4) states that riders of light motorcycles (legally one of the two categories of Mopeds) 13 and motorcycles must wear a helmet on the road. However this law is commonly ignored. 14 Figure 5 below shows that only 25.2% (187 in 842) of Moped riders wore helmets and just 15 2.7% (2 in 77) of motorcycle riders wore helmets. The proportion of Motorcyclists in 16 Shanghai is relatively low (only 5.3% of the total), compared to that of Moped riders (57.9%). 17 Also, motorcycles in Shanghai usually serve as an illegal mode of transit that only travel in 18 short distances. Thus, it may be that motorcyclists tend do not wear helmets for convenience 19 and business reasons. Interestingly, around 15% riders of bicycles and electric bicycles wear 20 helmets even though the law does not require it. 21

22

Figure 5 Proportion of Riders Wearing Helmets by Vehicle Category 23



1 In summary, Moped riders have several unique violation characteristics that separate 2 them from BT riders. The greatest difference is that a much higher proportion of Mopeds are 3 overloaded, and Moped violation patterns are much more complex. The next section will 4 consider the some of the key variables with which Moped violation patterns are associated. 5

3.2 Some Variables Associated with Violation Behaviors 6 The association of the Moped rider, vehicle characteristics, and the traffic environment to 7 violation behaviors (VBs) can be revealed by the use of Binary Logit models. The following 8 variables: rider demographics (gender and age), vehicle characteristics (type and power), 9 vehicle operation (direction), and traffic and road facility features of the intersection (split 10 green ratio of the approach, signal cycle, volume of riding vehicles at the approach, effective 11 width of the non-motorized vehicle lane at the approach, distance for non-motorized vehicles 12 passing through at the intersection, volumes of motorized vehicles, non-motorized vehicles 13 and pedestrians at crosswise lanes, ratio of saturation time of traffic flow on the crosswise 14 lanes and presence of traffic police) are all variables that may be predictive of any of the five 15 typical VBs (Red Light Running, Improper Lane Use, Improper Waiting, Improper Travel 16 Direction, and Vehicle Overloading). 17 Given that Red light Running VBs lead to the most frequent conflicts of any of the 18 five Moped VBs, we consider the Red Light Running Binary Logit model first. In the 12 19 intersections where video recordings were made, we found 16.6% of Moped riders violated 20 red lights, and of those, 46.6% were involved in conflicts. 21 Results of the Red Light Binary Logit model, Improper Waiting Violation Model, 22 and Improper Lane Use Violation Model are shown in Table 3 below. 23 24

Table 3 Variables Associated with Moped Red Light Running 25

Significant Variables Coef S.E P-value Constant

Signal cycle 1.473 -.007

.550

.003 ~.000 ~.000

Volume of RVs at approach -.001 .001 ~.000 Ratio of crossing traffic saturation time -2.662 .544 ~.000

Improper waiting before red light running 1.382 .255 ~.000 26 Moped Red Light Running VB Model 27 The Red Light Model results show that the all traffic environment factors (such as signal 28 cycle and the saturation time ratio) are significantly associated with red light violations while 29 none of the rider and vehicle characteristics factors were significantly associated with red 30 light violations. . Specifically, the short signal cycles and saturation time of crossing traffic 31 are associated with a great number of Moped riders’ red light running. This may occur 32 because when cycles are shorter there are more opportunities for Mopeds to follow one 33



another through red lights. In contrast, the longer cycles are associated with Moped riders 1 tending to stop and wait as more other Moped riders are doing so during the longer red light 2 time. The crossing traffic state is also an important variable associated with Moped red light 3 running. When the crossing traffic is less saturated, Moped riders are more inclined to violate 4 the red light. Also, Moped red light running increases as the volume of riding vehicles at 5 approach decreases, suggesting that riders would seize the chance to violate the red light 6 when fewer riding vehicles are present that may block their non-motorized lane. Improper 7 waiting behaviors are also predictive of red light running, as these two violations may share 8 the feature of riders’ being impatient. Table 4 shows the results of the improper waiting 9 Binary Logit model. 10 11

Table 4 Variables Associated with Moped Improper Waiting Violation 12


Moped operation direction (through) -.526 .620

.328

.235 .109

~.000

Non-motorized vehicle lane width at approach -.703 .133 ~.000

Improper lane use before improper waiting -1.381 .361 ~.000 13

Moped Improper Waiting VB Model 14 The VB of waiting for the green light at improper locations was committed by 12.1% (102) 15 riders according to the video records. This VB Binary Logit Model found that the direction of 16 Moped operation, and the width of the non-motorized vehicle lane are significantly 17 associated with this improper waiting violation. Through running Mopeds have comparably 18 higher VPs of improper waiting, and this could be linked to the shorter passing distance when 19 going straight. Also, when there is not enough room in the waiting zone area for riding 20 vehicles, Moped riders are also inclined to wait for the green light at an improper position (i.e. 21 beyond the stop line). Moreover, most Moped riders that wait at improper positions usually 22 use the non-motorized lane because motorized vehicles usually occupy motorized lanes when 23 the light is red. Table 5 presents the findings from the model for Improper Lane Use. 24 25

Table 5 Variables Associated with Moped Improper Lane Use Violation 26


Moped operation direction (through) Non-motorized vehicle lane width at approach

Motorized traffic flow at crosswise lanes Non-motorized traffic flow at crosswise lanes

Constant

1.310 -.834 -.839 -.009 -.001 1.310

.414

.201

.130

.001

.000

.414

`~.000 ~.000 ~.000 ~.000 ~.000 ~.000

27 28



Moped Improper Lane Use VB Model 1 Turning Mopeds are significant group for improper lane use VBs, as they ride out of the 2 non-motorized vehicle lane early to take a shorter route for left turns or to avoid the Mopeds 3 stopped for right turn. And, it is easier for Moped riders to use an improper lane when the 4 width of the non-motorized vehicle lane is insufficient and the motorized or non-motorized 5 traffic volume is small. Therefore, Moped riders are inclined to uses lane improperly when 6 traffic facilities are limited and interrupting traffic volume is small. 7 8 Other Moped VB Models 9 Red light running, Improper Waiting and Improper Lane Use violations of Mopeds are the 10 most severe and common VBs at urban signalized intersections. The improper travel direction 11 VB is not so common at signalized intersection areas in Shanghai, as only 51 riders (6.1%) 12 ran from an illegal direction. The model shows that the Moped improper travel direction 13 VB is only significantly related to non-motorized vehicle lane width at the entering approach 14 of an intersection (Coef=.382, P=.021). This VB occurs more often when the non-motorized 15 lane is wider, possibly due to the fewer potential conflicts when running from illegal direction. 16 Furthermore, 17.9% (151) of Moped riders commit the overloading VB. The Binary Logit 17 model shows that male riders (Coef=.596, P=.035) and gas powered light motorcycle riders 18 (B=.475, P=.013) are both more likely to overload than other riders. Better power 19 performance of Mopeds compared to BTs enable Moped riders to overload without 20 sacrificing too much performance. It should be noted that intersection traffic environment 21 factors have not been included in this model, as they do not affect the overloading decisions. 22 Therefore, according to the results of the violation behavior analysis of Moped riders 23 in Shanghai, the VP of typical Moped VBs varies significantly under different traffic 24 environment but is not significantly related to rider and vehicle characteristics. 25 26

3.3 Violations and Conflict Severity 27 The various conflict types identified above were classified into three severity levels, (slightly 28 interrupted, moderately interrupted, and full stop or sudden direction change, or, low severity, 29 moderate severity, and high severity). If we assign severity levels as follows: ‘slightly 30 interrupted’, 1; ‘moderately interrupted’, 2; ‘full stop or sudden direction change’ 3; the 31 weighted total percentage of Moped conflicts following each VB can be calculated. For the 32 five typical Moped VBs, the most likely occurring conflicts severity level are ‘Moderate 33 Severity’, while that of non-violation remains ‘Low Severity’. This is shown in Figure 6 34 below. 35 We found the severity of conflicts following a Moped violation was significantly 36 greater than the severity when there was no violation. We also found (when considering the 37 weighted total percentage of conflicts) that conflicts after VBs are more probable than 38 conflicts not preceded by VBs. Red Light Running violations result in the highest likelihood 39 of conflict, while the Improper Travel Direction results in the lowest. 40



1

Figure 6 Conflict Likelihood and Conflict Severity Percentages of the Five Typical Moped 2

Violation Behaviors 3

4 DISCUSSION AND CONCLUSIONS 4

Research to date has recently begun to examine the factors most closely associated with 5 Moped riders’ traffic behaviors in China. Previous studies have shown associations between 6 various Moped riders’ behaviors and variables such as gender, age, education, driving 7 technique, helmet use, vehicle state, traffic police presence (16), lane facilities (17), culture, 8 and education background (18). 9 This study analyzed violation behaviors (VBs) of Moped riders and VBs of bicycle 10 and tricycle (BT) riders, at 12 signalized intersections in Shanghai. The study found several 11 factors that can help us understand the antecedents and effects of Moped riders’ violation 12 behaviors (VBs). We found the traffic environment (i.e., traffic facilities, background traffic 13 and signal characters) to be more strongly associated with Moped VBs than either rider 14 characteristics (gender and age) or vehicle characteristics (vehicle type and vehicle power). 15 This means that different traffic environments affect the violation behaviors of Moped riders, 16 but riders’ behaviors do not differ much under the same traffic conditions. The implication is 17 that the traffic environment is the primary determinant of VBs. 18 We found that Mopeds riders commit more different types of VBs than BT riders. 19 Also, the severity of conflicts following a Moped violation was significantly greater than that 20 when there was no violation according to this study. This finding suggests it is likely that the 21



poor driving behaviors of Moped riders are placing others at risk. The regression model also 1 showed that red light violations are positively associated with improper waiting. Riders 2 running red lights depends to a large extent on the traffic conditions (red light violations 3 occur most often when apparently safe). The same is the case with other moped VBs in that it 4 is the traffic environment that limits or encourages improper behavior. For example, riders 5 tend to ride a little bit outside the non-motorized lane or wait outside the legal waiting zone 6 when the lane is tight. It should be noted that the data show fully 80.5% of light motorcyclists 7 are actually illegally sharing the non-motorized vehicle lanes with BTs. One likely reason 8 which could be is that these non-motorized lanes are much safer for mopeds than sharing the 9 lanes with cars and trucks. Thus, the law is not effectively restricting riders’ behaviors and, 10 therefore, should be revised. It was also shown that some electric bicycles riders wear 11 helmets even though they are not required to by law. This suggests that some riders have 12 strong safety concerns. 13 There continues to be a large gap in China in the understanding of Moped riders’ 14 behaviors, including their violation behaviors. Behavioral analysis based on video data can be 15 used effectively to examine the variables that underlie these behaviors and provide a stronger 16 basis for the development of countermeasures. Some possible countermeasures will be 17 considered in the next section. 18 19

5 RECOMMENDATIONS 20

Sound regulation could help to effectively control the use of Mopeds in China. Mopeds in 21 both Taiwan and in the US and are well regulated in their specifications and operations (19). 22 Also, in Taiwan, the policy is to encourage the construction of traffic facilities, such as 23 electrical charging stations, to support Moped use. (20). Regulations should serve as effective 24 guidance for Moped riders and need to be specific in encouraging or discouraging the use of 25 Mopeds. 26 Legislation should take into account the history, background and the prevailing 27 culture of Moped use to encourage better compliance with traffic law. It is recommended that 28 legislation be considered that covers Moped categorizations, pre-riding qualifications, 29 licensing, rules for Moped operation, and rules that specify precise enforcement protocols. 30 Specifically, it is recommended that the law allow light motorcycles to ride with 31 non-motorized vehicles because this would be safer for them then riding with cars and trucks. 32 However, the Moped vehicle specification should be strictly limited by law to ensure the 33 capabilities are do not exceed by too much the capabilities of bicycles and other vehicles that 34 use the non-motorized lanes. Also, the non-motorized lanes could be designed to have signs 35 and markings color matched to different types of Mopeds to help guide them to the left side 36 of the non-motorized lane. License tags with specific colors for identifying different types of 37 Mopeds and lane markings with corresponding colors could be considered. Traffic facilities 38 could also offer separate, safe and comfortable riding and designated stopping areas for 39



Mopeds. The improvement of traffic facilities should be combined with the driving education 1 that conveys the idea of facilities to riders. 2 While this paper has focused on Moped violations, previous crash research (especially 3 motorcycle crash research (24)) has shown that in collisions with cars and trucks, it is usually 4 the riders that are injured or killed, and it is often not the rider who is at fault in these 5 collisions. It also should be noted that even if the Moped violation behaviors we observed in 6 this study are found to frequently precede crashes, the extent to which they cause crashes is 7 still an open question. Future work will examine this and other questions more closely. 8

ACKNOWLEDGEMENTS 9

This study was jointly supported by the Chinese National Science Foundation 10 (51008230), the Chinese National Science and Technology Action Program for Road Traffic 11 Safety Study (2009BAG13A05), and the Scientific Research Foundation for the Returned 12 Overseas Chinese Scholars, State Education Ministry. 13 14

REFERENCES: 15

1. Weinert, J., Ogden, J., Sperling, D., and Burke A. The future of electric two-wheelers and electric 16 vehicles in China. Energy Policy, Vol. 36, 2008, pp. 2544–2555. 17

2. Dong, B. The Study of Characteristics of Electric Bicycle. Tongji University, 2008. 18

3. Cherrya, C. and Cerverob, R. Use characteristics and mode choice behavior of electric bike users 19 in China. Transport Policy, Vol. 14, 2007, pp. 247–257. 20

4. Traffic Safety Law of People’s Republic of China, 10th National People’s Congress, 2004. 21

5. Li, Y. Research on Traffic Accident and Injury Characteristics of Motorcycle. Third Military 22 Medical University, 2009 23

6. Chen, X., Han, H., Ye, J., Ma, S., and Xu, Y. Normalized Volume Measurement for 24 Non-Motorized Traffic Flow Mixed with Mopeds. Presented at Transportation Research Record: 25 Journal of the Transportation Research Board, No. 3266, Transportation Research Board of the 26 National Academies, Washington, D.C., 2011 27

7. Zhou, B., and Cheng, Y. Tricycle kills a passenger with scarf. Jfdaily, 2011-1-25, A1 04/05. 28

8. Jain, A., Meneze R., Nb, D., Kanchan, T., Gagan, and Jain, R. Two wheeler accidents on Indian 29 roads – a study from Mangalore, India. Journal of Forensic and Legal Medicine, Vol. 16, 2009, pp. 30 130–133. 31

9. Moskal, A., Martin, L., and Laumon, B. Risk factors for injury accidents among Moped and 32 motorcycle riders. Presented at Accident Analysis and Prevention, 2010. 33

10. Kasantikul, V., Ouelletb, J., Smithb, T., Sirathranont, J., and Panichabhongse. The role of alcohol 34 in Thailand motorcycle crashes. Accident Analysis and Prevention, Vol. 37, 2005, pp. 357–366. 35

11. Hurt, H.H., Thom, D.R., and Hancock, P.A. The effect of hand position on motorcycle brake 36 response time. Human Factors Society 28th Annual Meeting ‘New Frontiers for Science and 37 Technology’, San Antonio, Texas, 1984. 38



12. Kasantikul, V. Motorcycle Accident Cause Factors and Identification of Countermeasures in 1 Thailand Vol. I and II. KP Printing, Bangkok, 2002. 2

13. Li, Y., Wang, Z., Yi, Z., and Zhu, P. Research on Correlation of Traffic Accident with Driving 3 Behavior, Personality, Attitude towards Safety of Motorcycle Rider. Psychological Science, Vol. 4 31, 2008, pp. 487–489 5

14. Wang, Y., Ma, G., and Tang, G.. Analysis on Influence of Electric Bike to Urban Traffic Safety. 6 Communication Standardization, Vol. 138, 2005, pp. 27-29. 7

15. Tiwari, G., Mohan, D., and Fazio, J. Conflict analysis for prediction of fatal crash locations. 8 Accident Analysis & Prevention, Vol. 30, 1998, pp. 207-215. 9

16. Dandona, R., Kumar, A., and Dandona, L. Risky behavior of drivers of motorized two wheeled 10 vehicles in India. Journal of Safety Research, Vol. 37, 2006, pp. 149–158. 11

17. Yang, L., and Zhang, C. Research on riding behavior of bicycle drivers. Highway traffic and 12 safety, Vol. 6, 2006. 13

18. Njå, O., and Nesvåg, S. Traffic behavior among adolescents using Mopeds and light motorcycles. 14 Journal of Safety Research, Vol. 38, 2007, pp. 481–492. 15

19. Ding, Z. Research on the Regulation of Electric Bicycle Safety in the US. Overseas Inspection, 16 Vol. 9, 2009, pp. 61-64. 17

20. Tso, C. and Chang, S. A viable niche market—fuel cell scooters in Taiwan. International Journal 18 of Hydrogen Energy, Vol. 28, 2003, pp. 757–762. 19

21. Goldenbeld, C., Twisk, D., and Craen, S. Short and long term effects of Moped rider training. 20 Transportation Research Part F, Vol. 7, 2004, pp. 01–16. 21

22. Institute of Transportation Engineers, and Federal Highway Administration. Behavioral Analysis 22 as Part of Pedestrian Accident Monitoring. ITE 2008 Technical Conference and Exhibit, 2008. 23

23. Lahey, B. Psychology: An Introduction (9th edition). Shanghai People’s Press, Shanghai, 2010. 24

24. Hurt, H.H., Ouellet, J.V., and Thom, D.R. Motorcycle Accident Cause Factors and Identification 25 of Countermeasures, Volume 1: Technical Report, Traffic Safety Center, University of Southern 26 California, Los Angeles, California 90007, Contract No. DOT HS-5-01160, January 1981 (Final 27 Report). 28


Moped Rider Violation Behaviors and Moped Safety in · PDF file2 Moped Rider Violation...

Documents

Transcript of Moped Rider Violation Behaviors and Moped Safety in · PDF file2 Moped Rider Violation...