We extracted 463 cases of crossing collisions of mopeds and larger vehicles from studies of accidents made by ITARDA from 1993 to 2001. Included among the primary parties were some who had died and others who could not recall the events of the accidents clearly, making it impossible to learn the human factors of the accident. As a result, we were able to obtain data on errors from 832 people (409 Parties A and 423 Parties B). "Party A" means the party committing the error that caused the accident, while "Party B" refers to the other party. These terms are not the same as the "primary party" and "secondary party" used by the police. As explained earlier, one party often makes a variety of errors at different points in time. When all of these errors are totaled, the 409 Parties A committed 1286 errors, while the 423 Parties B committed 927 errors. (1) Outline of Errors   As already touched upon, the total number of errors that the 832 Parties A and B made was 2213. Figure 8 shows the numbers of errors per person. The Parties A committed 3.1 errors per person, while the Parties B committed 2.2 errors per person. In other words, it is clear that even the Party B side had a good chance of avoiding the accident.   The difference between the Parties A and Parties B lies in the shares of the types of errors. The percentage of errors made by the Parties A in the recognition stage was greater than that of Party B by approximately 20 percentage points. Conversely, the percentage of errors in the judgment/prediction stage was 20 percentage points lower. Errors in the operation/action stage were extremely rare and are hence omitted.   The results of the analysis for the Parties A and Parties B are explained below. (2) Objects of Recognition Errors   Figure 9 classifies all 1560 occurrences of recognition errors by type of object. Of these cases, 81% were of vehicles in intersections, 9% were of traffic lights, and 5% were of stop signs. An examination of these three objects is believed sufficient for the discussion of the causes of crossing collisions. In the next section, we will explain the situations behind errors regarding these three objects. Note that "vehicles in intersections" refers to vehicles that are driving through intersections or other crossing roads.

(3) Situation Behind Errors by Object   [Errors Regarding Vehicles in Intersections] a) Factors in Recognition Error (Reasons)   Figure 10 shows the factors involved in recognition errors of vehicles in intersections. The vertical axis shows the types of factors involved in recognition errors. In addition, the subcategories "could not see" and "did not pay attention" are displayed in the legend.   "Did not pay attention" accounted for over 42% (535 cases) of recognition errors. As displayed in the legend, recognition error factors are "green light or right of way", "already confirmed", and "little traffic". "Lowered concentration" and "paid attention to a different object" are factors where the driver had no intention of looking at the object. Hence, these factors are considered to be ones where the driver "did not look". While we use the term recognition error, many of these errors are results of judging for some reason that the object "does not have to be looked at". This "judgment for some reason" can be said to be a "habit" caused by past experience where the driver had not been in any accidents before.   In contrast, errors classified as "could not see" accounted for 34% (428 cases) of recognition errors. The majority of these are "concealment by buildings, structures, trees", while 57 cases were "concealment by other cars". Elimination of those road environments with poor visibility that increase the burden on drivers and cause errors is desirable, however it is still important when recognizing that visibility is hindered that there may be something dangerous behind the obstruction.

b) Judgment/prediction errors   What are most drivers thinking when they successfully recognize a vehicle in an intersection? In the final analysis, over 90% drivers believed that for some reason "they could proceed first". We have summarized the factors (reasons) leading to them judging "they could proceed first" in Figure 11. The main ones were "seen by the other driver also" and "green light or right of way". Almost always these factors were found among Parties B. [Errors Involving Traffic Lights and Stop Signs] a) Recognition Errors   While not shown in the figures, approximately half of the factors involved in recognition errors of traffic lights and stop signs are classified as "lowered concentration" such as "daydreaming" and "carelessness". However, about 13% of traffic light related cases fell into the category "already confirmed". These are recognition errors resulting from the traffic light being green when initially checked, but not being reconfirmed when the vehicle was entering intersection. These are error factors that occur when viewing a traffic light constantly changing between red, green, and yellow. b) Judgment/prediction errors   Next, we will explain the judgment/prediction errors that are made when recognizing a stop sign. Over 85% of these errors were those in which the driver judged that it was acceptable to ignore the stop sign. Figure 12 shows the factors that are involved in unsuitable judgment/prediction. The majority of the factors were ones where the driver "confirmed that there were no other vehicles". Drivers have a strong tendency to ignore stop signs if they believe that there are no other vehicles on the road.   Even if really confirming that there was no one on the road, it wouldn't be so bad, but usually the judgments are made based on inadequate confirmation while just slowing down. This leads into accidents. Generally, stop signs are installed at intersections with poor visibility, so it is crucial to stop, as often stressed, to stop twice and check for the presence of other vehicles.   Figure 13 details the judgment/prediction errors involved in recognition of a traffic light and their error factors. In 50 (49%) of the cases, the drivers did not "pay attention". Of those, 45 "paid no attention" and proceeded forward because the oncoming traffic light changed from red to green and collided with a vehicle that ignored the traffic light at the intersection. These made up approximately 10% of accidents that were targeted in this analysis, and 35% of those accidents limited to traffic lights at intersections.   This is not a critique of immediately accelerating when a light turns green, but if drivers were to take a deep breath before acting, approximately 10% of accidents would not have occurred. This is called defensive driving.

UP

Institute for Traffic Accident Research and Data Analysis (ITARDA)