Cover Image

Not for Sale



View/Hide Left Panel
Click for next page ( 80


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 79
A P P E N D I X A Project 2 Data Dictionary Motivation and intent were important criteria. The motivations 2 = Snow. (This category included sleet. Several cues helped behind the driver's actions, when able to be clearly determined, indicate if the precipitation was in fact snow; snow tended to weighed more heavily than actual behaviors. Events were coded be larger and fall more slowly than rain, it looked like white according to judgment with respect to why the alert went off-- flurries and was also present on the ground, reinforcing the as though one was looking through the eyes of the driver. classification as snow. Also, precipitation that occurred in December through February was assumed to be snow, not Time of Day rain. Snow could be coded in other months, but the assump- tion that the precipitation was snow was not as strong.) Dusk was difficult to classify as either day or night; this classi- fication was subjective. Events were coded to reflect whichever the clip was most similar to, day or night. Location of Eyes at Time of the Alert 0 = Day. This category was coded at the actual time of the alert, when 1 = Night. the radar display showed 100 for the alert level. Eye location was coded by what the reviewers could see of the driver's eyes Road Condition at the time of the alert, even if they could not see the eyes preceding the alert. Reviewers coded the location of the Glare and reflection helped determine whether the road was driver's eyes even if they could see only one eye because it was dry or wet. assumed that the driver's eyes moved in parallel. Because of the absence of an eye-tracking camera and the limitations 0 = Dry. of the face camera, some ambiguity often existed about where 1 = Wet. (Any moisture on the road led to the classification the drivers were looking. Reviewers needed to be confident in as wet; there did not need to be standing water. The road the location of the driver's eyes in order to code as a specific was classified as wet if it was wet from snow but not snow covered.) location. In many instances, the reviewers were confident that 2 = Snow covered (Snow covered would have included ice the driver's eyes were not looking forward but could not tell covered if it was observed, but it was never observed. If any specifically where the eyes were looking. These instances were portion of the road, including turn lanes, was covered in coded as 8s. One such example is when the driver appeared to snow, then the classification was snow covered.) be looking at the camera. In this situation, it was difficult to determine if the driver was looking at the camera intention- ally, glancing out the corner, or looking slightly out the left Precipitation window; therefore, it was coded as an 8. The determination of Spots on the windshield or wiper activity helped determine if whether glances were still forward or if they were away was also there was in fact precipitation. difficult and subjective. Reviewers agreed on an area or "box" that they considered to be looking forward; this allowed for 0 = None. slight glances but even many scans across the forward scene 1 = Rain. (Light rain and drizzle were classified as rain, as were considered glances away. This process defined "looking were downpours.) forward" narrowly and essentially meant straight forward. 79

OCR for page 79
80 Glances toward the right of the forward scene, the right area of resents the time between the change in gaze and the alert. Time the windshield, were glances away and were coded as 8s. was recorded in 10ths of seconds. If the driver was always look- ing forward, then the time from the last nonforward glance was 0 = Looking forward at forward scene. (Looking forward left null because that category was not applicable. If the driver included looking at the head-up display [HUD].) was looking away 0.1 s before the alert and then was looking 1 = Left outside mirror or window. forward at the time of the alert, the time from the last nonfor- 2 = Looking over left shoulder. (The driver's gaze needed ward glance was recorded as 0. If the driver's eyes were not to look over the driver's shoulder, although the driver's visible, typically because of glare, for any portion of the clip, chin did not necessarily need to cross over the driver's the location was coded as a 7 because one could not be certain shoulder.) there was not a glance away. The only exception to this rule was 3 = Right outside mirror or window. when the reviewers could not see the driver's eyes and then the 4 = Looking over right shoulder. (The driver's gaze needed eyes became visible so that the reviewers could see the eyes and to look over the driver's shoulder, although the driver's chin there was a glance away before the alert. This situation negates did not necessarily need to cross over the driver's shoulder.) the fact that the reviewers could not see the eyes at the begin- 5 = Head down, looking at instrument panel or lap area. ning of the clip, because there was a nonforward glance after (Looking at the HUD was not considered part of the instru- the portion during which the eyes were unclassifiable. If the ment panel.) eyes were then unclassifiable again, before the alert but after the 6 = Head down, looking at center stack console area. (Con- glance, the eyes were coded as a 7 because the reviewers could sole means the area in which the stereo, thermostat, and not be certain what happened during that portion of the clip. clock are located.) If the location of one eye could be determined but not that of 7 = Driver wearing sunglasses or glasses with glare. (Glare the other eye, location was still coded. Reviewers were confident prohibited the ability to classify where the eyes were look- in coding eye position when only one eye could be seen because eyes normally move in parallel. If the driver's eyes glanced away ing. In some instances, drivers were wearing sunglasses, before the alert and were in transition at the time of the alert, but the reviewers thought that they could confidently iden- the last nonforward glance code reflected where they were look- tify the location of the drivers' eyes. In these instances, eye ing at the time of the alert, not where they had previously been location was recorded.) looking. For more details on eye location, see the information 8 = Cannot accurately evaluate eye location. (An 8 was cho- on eye location at the time of the alert. The criteria for classify- sen when the reviewer was unsure of the eye position or ing a glance as a specific location are the same as the criteria for classification within a reasonable level of confidence but eye location at the time of the alert. not because of glasses. Typically, the reviewer could see the actual eye but could not determine where the gaze was 0 = Always looking forward at the forward scene. (Looking directed. Eyes in transition were often coded as 8 because forward includes looking at the HUD.) it was unclear where the driver's gaze was at that particular 1 = Left outside mirror or window. moment.) 2 = Looking over left shoulder. 9 = Other. (For example, the driver may clearly be look- 3 = Right outside mirror or window. ing at the passenger side floor. When a glance was coded 4 = Looking over right shoulder. as other, the location was noted in the notes section. The 5 = Head down, looking at instrument panel or lap area. most common position recorded as other was the rearview 6 = Head down, looking at center stack console area. (Con- mirror.) sole means the area in which the stereo, thermostat, and clock are located.) 7 = Driver wearing sunglasses or glasses with glare. (Glare Location of Eyes During the prohibited the ability to classify where the eyes are looking.) Last Nonforward Glance 8 = Cannot accurately evaluate eye location. (An 8 was cho- and Time from the Last sen when the reviewer was unsure of the eye position or Nonforward Glance classification within a reasonable level of confidence but If the driver's eyes were on the forward scene at the moment of not because of glasses. Typically, the reviewer could see the the alert but had looked away during some portion of the clip actual eye but could not determine where the gaze was previous to the alert, this location was recorded. The reviewers directed. Eyes in transition were often coded as 8 because also recorded the amount of time between when the driver's it was unclear where the driver's gaze was at that particular gaze began to return to the forward scene and the moment of moment.) the alert, according to the radar display showing alert level 100. 9 = Other. (For example, the driver might clearly be look- We did not count the actual moment of the alert; the time rep- ing at the passenger side floor. When a glance was coded

OCR for page 79
81 as other, the location was noted in the notes section. The the code was null (not applicable). If reviewers were not sure of most common position recorded as other was the rearview the location of the driver's eyes, then the time to visual response mirror.) was left as null. The time to visual response was recorded for Week 1, even though there was no alert to which to respond. The rationale for coding this was that a baseline would provide Eyes on Task at Time of the Alert an idea of what a normal time to visual response was compared 0 = No. (The classification of no was used only when the with the time to response with an alert. reviewer could confidently determine that the driver's eyes were off the task of driving at the time of the alert [e.g., the 0 = Looked in response. (The driver initiated a look in driver was looking at a friend or the stereo system].) response to the alert within 1.0 s. Glances qualified as a 1 = Yes. (The classification of yes does not mean looking look in response.) forward; it means that the driver's eyes were on the task of 1 = Did not look in response to alert. (The driver did not driving.) look within 1.0 s of the alert.) 2 = Cannot determine. (For instance, the driver was wear- 2 = NA. (This option was always used for Week 1 because ing glasses with glare or reviewers could not see the dri- no alert occurred during that week; thus, this category could ver's eyes for some other reason. This classification was also not be coded. This option was also selected when the driver used when reviewers could not tell if the eye location was was already looking forward at the time of the alert; this on task [e.g., the driver was looking out the window, but it category was not applicable.) was unclear whether the driver was looking at traffic or at 3 = Cannot tell. (Because the driver was wearing sunglasses a fancy building that was distracting the driver's attention]. or other glasses with glare, reviewers could not tell where In any case, reviewers did not know whether the driver was the driver's eyes were looking.) on task.) Visual Occlusion Eyes in Transition Occlusion was coded with regard to the driver as well as to To classify the driver's eyes as in transition, they must have the reviewers. For instance, heavy rain or bright sun might been in transition at the time of the alert and must have started have occluded the scene for both parties, whereas blurry the transition at least 0.1 s before the alert. The eyes could not video occluded the scene only for reviewers. The occlusion be at the beginning of a transition or the end of one; they must did not necessarily have to impact the reviewers' ability to have been in the transition at the time of the alert. code the scene. 0 = No. 0 = None. 1 = Yes, toward forward scene. 1 = Sun or headlight glare. (This classification includes 2 = Yes, away from forward scene. when the scene was whitewashed from the sun. Only head- 3 = Cannot tell. (Cannot tell was selected when the driver light glare was included in this section; taillight glare was was wearing sunglasses or reviewers could not see the driver's coded as other.) eyes for some other reason; therefore, it was uncertain 2 = Other, specified in notes section. (The most common whether they were in transition.) entry was taillight glare.) Visual Response to Alert and Startle Response Time to Visual Response Coding the startle response was subjective, and the classifica- Reviewers coded the time that it took the driver to initiate a tion as such was often hotly debated. The driver had to be vis- visual response to the alert by filling in the number of 10ths of ibly rattled. The driver's startle was observed by body response, a second the response took. The time counted was the time dialogue, or both. between the alert and when the look was initiated, not includ- Cursing was not sufficient to be coded as startle, because it ing the moment of the alert or the moment of response. If the might have resulted from anger or frustration, not startle. response was initiated within 1.0 s, then the driver was consid- This category tried to capture startle either to the situation or ered to have looked in response to the alert. The amount of to the alert. time it took to look in response was always recorded for appli- cable situations, even if this was greater than 1.0 s. If the driver 0 = No. was already looking at the road and continued to look forward, 1 = Yes.

OCR for page 79
82 Steering in Response that were relevant. Lane shifts were also considered shifts in traffic patterns, such as lane shifts in construction areas.) 0 = No steering in response to alert. (Small, jerky reactions or slight wiggling in response to the alert or to the situation was classified as a 0 and not considered steering.) Secondary Driving Behaviors 1 = Driver steered partially or fully in response to the alert. Audio was used to assist in coding whenever possible. For (Steering, for review purposes, was an evasive maneuver in instance, reviewers may have heard the radio station change an attempt to prevent striking a vehicle; thus there must and seen the driver look at the console; this would indicate in- have been a significant amount of steering.) car system use. The default for nondriving behaviors was none, coded as 0. Hand Location at Time of Alert Because both hands were not often visible, reviewers coded Cell Phone what could confidently be inferred from the scene. At times 10 = Conversation, in use. (Conversation could be coded playing the video farther helped determine what was ambigu- for listening, talking, or both while using the cell phone.) ous in a still frame at the time of the alert. For instance, at 11 = Reaching for phone. (This classification refers to when the time of the alert there may have been a small blur near the the driver reached for the handheld phone to speak on that steering wheel. On continuation of the video the blur may have phone. If the driver reached for the phone simply to answer moved and come into view as a hand. the phone and talk on the headset the driver was wearing, the classification was other. Simply answering the phone 0 = Cannot see the position of either hand or cannot deter- involves far less physical activity by the driver than reach- mine the position of either hand. (Reviewers coded 0 if a ing for the phone and holding it during a conversation.) hand could be seen but they could not tell if it was on the 12 = Dialing phone. wheel.) 1 = At least one hand on the steering wheel. (Reviewers used this code when the position of one hand could not be Headset, Hands-Free Phone determined but they could see that at least one hand was 20 = Conversation. (This classification was selected when on the steering wheel.) reviewers could tell that the driver was in a conversation.) 2 = Both hands on the steering wheel. 21 = Reaching for headset. 3 = At least one hand off the steering wheel. (This code was 22 = Unsure of activity level. (The driver was wearing a used when the position of one hand could not be determined headset, but it was not clear whether the headset was in use. but at least one hand was clearly off the steering wheel.) The driver may have been listening to someone or wearing 4 = One hand on, one hand off the steering wheel. (A 4 was the headset in case there was an incoming call.) classified when reviewers could clearly see both hands and one was on the wheel and the other was off.) Eating 5 = Both hands off the steering wheel. (A 5 was classified when reviewers could clearly see both hands and both were 30 = Highly involved. (High involvement includes such off the wheel.) things as eating a burger or unwrapping food.) 31 = Low involvement. (Low involvement includes such things as eating candy or grabbing chips.) Road Geometry 0 = Straight. Drinking 1 = Curve. (A curve must be of some substance to be con- sidered a curve.) 40 = Highly involved. (High involvement includes situa- 2 = Approaching curve. (The classification of approaching a tions in which the driver was trying to open a straw or bot- curve constituted situations in which the driver was almost tle or was blowing on a hot drink.) in a curve, not when there was simply a curve in the distance.) 41 = Low involvement. (Low involvement includes situa- 3 = Lane shift. (Road geometry was classified as a lane shift tions in which the driver was sipping a drink or drinking when there was a change in the lane structure, for instance, without looking.) when a new lane was created. If the new lane was a turn lane, 50 = Conversation. (The driver and someone in the car it was not classified as a lane shift, because the scenario were carrying on a conversation. The driver can be listen- should have covered the fact that there was a turn lane if ing during the clip, talking during the clip, or both.)

OCR for page 79
83 60 = In-car system use. (The driver was actively adjusting the driver may follow it, is more in line with that of a car than something. For example, the driver was not just listening with that of an SUV. to the stereo but also adjusting it. The car lighter was coded under the smoking section.) 0 = Vehicle in path--car. 1 = Vehicle in path--pickup, van, or SUV. 2 = Vehicle in path--other (e.g., motorcycle, semitrailer, Smoking commercial vehicle). 70 = Lighting. (This classification includes the in-car lighter.) 3 = Vehicle out of path--car. 71 = Reaching for cigarettes or lighter. (This classification 4 = Vehicle out of path--pickup, van, or SUV. includes the in-car lighter.) 5 = Vehicle out of path--other (e.g., motorcycle, semi- 72 = Smoking. trailer, commercial vehicle). 6 = Construction. (This includes all equipment associated with construction [e.g., barrels, cones, and construction Grooming vehicles].) 80 = Highly involved. (High involvement includes applying 7 = Discrete roadside object. (This classification includes makeup or brushing hair.) signposts, light poles, trees, fire hydrants, and mailboxes.) 81 = Low involvement. (Low involvement includes scratch- 8 = Overhead items. (This classification includes such ing or running one's fingers through his or her hair.) items as overhead signs and bridges.) 90 = Other/multiple behaviors, specified in notes section. 9 = Bridge support. (Such behaviors might include whistling or classifications 10 = Guardrail/Jersey barrier. that reviewers were unsure of [e.g., if the driver's lips were 11 = Other, to be specified in notes section. moving but there was no audio, the behavior might be singing or conversation].) Target Stationary or Moving This category was coded using both visual cues and radar data. Alert Classifications For vehicles that appeared to be either stopped or slowly mov- ing, visual cues were used if the cues were clear--for example, 1 = False alarm. (Two reasons were given for the classifica- lane markings--but otherwise, radar data was used to deter- tion of an event as a false alarm. The first was that the tar- mine the classification. get was out of the lane throughout the episode or it was off the roadway, including both vehicles and stationary objects. 1 = Stationary. (The target must have had a velocity of less The second was that the kinematics of the scenario did not than or equal to 1.3 m/s to be classified as stationary.) make sense [e.g., when there was a lead vehicle deceleration 2 = Moving. error resulting in an alert].) 3 = Stationary potential threat. (The classification of 3 2 = True/nuisance alert. (The target had to be a vehicle that was made when the target was a vehicle that was stopped was in the driver's path for at least a portion of the episode. [velocity of less than or equal to 1.3 m/s] but had the The event may have been viewed by the driver as either a potential to move at any moment [e.g., stopped cars with needed alert or a nuisance alert.) drivers in them].) 3 = Instigated alert. (An alert was classified as instigated if the driver deliberately drove in such a way as to provoke an alert. This does not apply to when the driver was in adap- Forward Conflict Scenario tive cruise control [ACC] and was trying to see if the system The same scenarios are used for classifying supplementary would brake sufficiently; this was testing the ACC system scenarios as needed. rather than the forward crash warning [FCW] system.) The supplementary scenario column also has the option of 0 = none. Supplementary scenarios are available in case there Target is another possible scenario or if the situation resembles another scenario that may be of interest to people working with that A vehicle was considered in path if even a small portion--for specific type of scenario. example, a rear bumper--of the lead vehicle remained in the driver's lane at the time of the alert. Vehicles with both sport- Out of Host's Path utility vehicle (SUV) and car characteristics, a small SUV, were classified as a car. The classification as a car for these 100 = False alarm. (Two reasons were given for the classi- SUVs was because the body of the vehicle, in respect to how fication of an event as a false alarm. The first was that the

OCR for page 79
84 target was out of the lane throughout the entire episode or signal was in view. The event was coded as 240 even if the lead it was off the roadway, including both vehicles and station- vehicle had not stopped by the time of the alert, as long as the ary objects. The second was that the kinematics of the lead stopped within 12.5 s of the alert [the regular viewing scenario did not make sense [e.g., when there was a lead time plus a 10-s continuation]. For coding purposes, stop vehicle deceleration error resulting in an alert].) means that the lead came close to 0 mph [less than or equal to 1.3 m/s].) 250 = Host follows a lead vehicle that decelerates to a In Host's Path predictable stop. (This scenario occurred when there was a The following scenarios were initiated by the host vehicle. traffic light, stop sign, visibly stopped traffic, or the car in Note, however, that this and the following subcategories front of the lead had its brake lights on. These cues made apply to the stereotypes of scenarios and may not apply to it so that the host could logically anticipate that the lead all cases. Nevertheless, these assumptions were used during would stop. The event could be coded as 250 even if the some analyses. lead had not stopped by the time of the alert, as long as it had stopped within 12.5 s of the alert [the regular viewing time 200 = Host tailgating. (Tailgating was coded even if the lead plus a 10-s continuation]. For this scenario, a stop means vehicle was on an exit or entrance ramp. The criterion for coming close to 0 mph [less than or equal to 1.3 m/s].) tailgating was a headway of 0.8 s or less. If the host was using 260 = Host follows a lead vehicle that decelerates in an ACC and the headway matched the criterion for tailgating, unpredictable manner but does not stop. (For this scenario it was not considered tailgating, because the system rather the brake lights on the lead vehicle had to be visible or the than the driver was controlling the headway.) lead was noticeably slowing even if the brake lights could 210 = Host approaches an accelerating vehicle. (This typi- not be seen but the reason for slowing was not visible or cally occurred when the host misjudged the lead vehicle's not predictable [e.g., a cat crossing the street]. The classifi- acceleration and the host accelerated too fast on a vehicle cation as 260 was the default over a 270.) that was accelerating as well or if the host was approaching 270 = Host follows a lead vehicle that decelerates in a pre- a vehicle as a traffic light turned green.) dictable manner but does not stop. (For this scenario the brake lights on the lead had to be visible or the lead was The following scenarios played out naturally with neither noticeably slowing even if the brake lights could not be the host nor the lead vehicle changing accelerations. seen. The code of 270 was selected only if it was clear from the cues available why the lead needed to decelerate [e.g., a 220 = Host approaches slower vehicle that is traveling at slow moving piece of farm equipment was ahead or the car constant speed. (The criteria for this classification were as in front of the lead had its brake lights on].) follows: no brake lights were visible; lead vehicle was always going slower than the host while the target was detected by Transitional Host Path: One or ACAS; and the target did not appear to be decelerating, Both Vehicles Change Lanes either visibly or using the radar data. Slight fluctuations in speed are normal, but the overall speed must have been The following scenarios were initiated by the host vehicle. fairly constant. Typically, the lead vehicle was in the dis- tance at the beginning of the clip and then moved into view 300 = Host cuts behind a lead vehicle. (The code 300 was used as the host gained on it.) when the host cut into another lane closely behind the lead 230 = Host approaches lead vehicle that is already stopped. vehicle. This maneuver could not be part of a two-lane pass. (The lead vehicle must have been traveling less than 1.3 m/s See the description of a two-lane pass for more details.) during all portions of the clip in which it was acquired as a 310 = Host performs two-lane pass in order to pass. (This target. Please see the notes on Target Stationary or Moving scenario involves the host's cutting behind the lead in the for more details on how the determination of a target as sta- process of making a two-lane pass maneuver: the host tionary or moving was made.) crossed the middle lane in order to enter two lanes over from the host's lane of origin. If two alerts occurred for The following scenarios were initiated by the lead vehicle. the same smooth transition during a two-lane pass, review- ers coded the first alert as 310 and did not code the second 240 = Host follows a lead vehicle that decelerates to an un- alert in the series. Reviewers commented on the scenario predictable stop. (This scenario was classified as such when in the notes section, labeling the second alert as such in the traffic stopped but the host driver may have thought the notes section. A two-lane pass does not require two traffic would just slow and not stop; no stop sign or traffic alerts.)

OCR for page 79
85 315 = Host performs two-lane pass to exit or to make a turn in front of the host. This occurred when the lead was in a that is carried out within the time of the clip. (The scenario lane parallel to the host's and then cut in or merged close was classified as such when the host cut behind a lead in to the front of the host's vehicle.) the process of making a two-lane pass maneuver: the host 365 = Lead vehicle is forced to merge in front of host. (An crossed the middle lane in order to enter a lane two lanes event was classified as 365 when the lead needed to merge over from the host's lane of origin. If two alerts occurred into the host's lane because an entrance ramp was ending for the same smooth transition during a two-lane pass, or a lane was ending for another reason.) reviewers coded the first alert as 315 and did not code the 370 = Lead executes a two-lane pass. (This scenario was second alert in the series. Reviewers commented on the coded when the lead passed from a lane parallel to the host's, scenario in the notes section, labeling the second alert across the host's path, and then over to the parallel lane on as such in the notes section. A two-lane pass does not require the other side of the host's car. The lead was only in the host's two alerts.) path momentarily.) 320 = Host changes lanes to pass lead vehicle. (The tar- 380 = Lead vehicle in adjacent lane weaves or encroaches get for this alert type was the vehicle in the host's original unintentionally/unknowingly into host's lane. (The classi- lane. For this event, a pass was coded even if the host had fication of an event as 380 refers to events in which the lead not passed at the time of the alert, as long as the host was act- entered the host's lane unintentionally and momentarily. ing like a pass was planned; for instance, the host was The brief entry into the host's lane by the lead vehicle caused checking the mirrors and accelerating. If the pass was the alert.) aborted and not carried out during the entirety of the 390 = A vehicle crossing the host's roadway travels straight extended clip [10 additional seconds], reviewers classified across host's path. (The scenario is characterized by a vehicle the event as a pass and marked the scenario as aborted. driving straight across the host's path. The target vehicle did Reviewers added notes when necessary. If the host was not remain in the host's path; typically, the radar hit the passing a vehicle that was waiting to turn, reviewers coded side of the crossing vehicle, and the intersecting paths were whichever event seemed to be the reason for the alert and perpendicular to each other.) the other event could be chosen in the supplementary 400 = A vehicle makes a left turn across the host's path in scenario section.) order to travel in a direction other than that of the host. (In 330 = Host enters turn lane or other dedicated lane (e.g., exit this scenario the crossing vehicle turned left either from a lane) while approaching a lead vehicle in the new lane. (In perpendicular street or from a parallel lane traveling in the this case, the target was the vehicle in the host's new lane.) opposite direction of the host's lane. If the vehicle turned 340 = Host enters a turn lane or other dedicated lane (e.g., from a side street, it crosses the host's path and then con- exit lane) while approaching lead vehicle in original travel tinues into a parallel lane traveling in the opposite direction. lane. (In this case, the target was the vehicle in the host's If the vehicle turned left from a parallel lane, it crossed the original lane.) host's path and continued onto a perpendicular street. The 350 = Host weaves in order to avoid an obstacle but does not vehicle crossed the radar path with primarily a perpendi- completely change lanes. (An event was coded as 350 if the cular angle, but the angle may be more steeply tilted than host did not completely change lanes in the process of avoid- when the vehicle was simply crossing straight across the ing an obstacle. This was not a planned maneuver; there was host's path in other scenarios.) no turn signal or other indications of a lane change until the 410 = A vehicle entering the host's roadway crosses host's last moment--an evasive reaction.) path and moves into a lane parallel to the host's lane in the 355 = Host weaves in order to avoid an obstacle and com- same direction. pletely changes lanes. (An event was coded as 355 if the 420 = A vehicle pulls out from a side street or driveway and host changed lanes completely in the process of avoiding an pulls in front of host and into the host's lane. (This sce- obstacle. The motivation for the lane change has to have nario occurred when the lead started out perpendicular to been to avoid something in the original lane. This was not the host car and turned into and in front of host's car.) a planned maneuver; there was no turn signal or other indi- 430 = Lead changes lanes out of host's lane. (This sce- cations of a lane change until the last moment--an evasive nario developed when the lead departed the host's lane reaction.) but the situation was not covered by another described scenario. One instance of this situation is when the host The following scenarios were initiated by the lead vehicle. could logically anticipate that the lead vehicle would change lanes because of the lead's turn signal or another indica- 360 = Lead vehicle changes lanes and cuts in front of host. tion and, therefore, the host gained on the lead, sounding (The main precipitant of this scenario was the lead cutting an alert.)

OCR for page 79
86 440 = Lead leaves host's lane to enter turn lane or other Notes dedicated lane (e.g., exit ramps and turn lanes). 450 = Lead turns left from host's travel lane. (The target in A notes section recorded any unusual events or ambiguous this conflict is the turning car. The lead must have begun situations not covered by categories for a particular question. the turn, even slightly, for the scenario to be coded as 450. This section also contains general notes on the clip if there If the lead's turn signal was on but the turn had not yet was anything significant taking place that was not adequately been initiated, the event was coded as a predictable decel- covered by the coding process. See section Annex A-1 below eration and given a supplementary scenario of 450.) for further details. 460 = Lead turns right from host's travel lane. (The target in this conflict is the turning car. The lead must have begun Annex A-1 the turn, even slightly, for the scenario to be coded as 460. If the lead's turn signal was on but the turn had not yet The following are examples of items that are captured in the been initiated, the event was coded as a predictable decel- notes section, although other, unforeseen events are also noted. eration and given a supplementary scenario of 460.) Visual Occlusion Scenario Completion Rear taillights, glare from rain and wetness on the road, blurry 0 = Completed. video, dirty windshield, temporary incapacitation, sneezing, 1 = Aborted. flying debris, faulty wiper or defroster, and an object in or over the eyes. Supplementary Scenario Nondriving Behaviors This category allowed reviewers to select a second scenario to which the situation could be attributed. The supplementary Whistling, two or more behaviors, if there is no audio and the scenario could also be a scenario that preceded or followed the driver is clearly talking or singing but reviewers could not tell imminent scenario but may have contributed to the develop- which, attempting to avoid an insect in the car, adjusting mir- ment of the actual scenario. This category was designed to rors, reading a map, reading other materials, checking a watch, be a reference list for people interested in scenarios that could or yawning. have been classified in other ways. The category also allowed reviewers to indicate two scenarios when the primary scenario Target was difficult to determine. If reviewers did not indicate a sup- plementary scenario, a 0 was entered for none. Shadows and embankments.