• Federal Highway Administration >
• Publications >
• Research Publications >
• 06033 >
• 03scenario4.Cfm >
• Task Analysis of Intersection Driving Scenarios: Information Processing Bottlenecks

Task Analysis of Intersection Driving Scenarios: Information Processing Bottlenecks

PDF files can be viewed with the Acrobat® Reader®

Scenario 4–Straight on Green Light

Description

This scenario involves the subject vehicle going straight through the intersection on a green light after changing lanes to avoid a vehicle that is stopping to turn in the left lane. Figure 29 shows the scenario diagram and provides additional details regarding the scenario. Briefly described, this scenario involves the driver approaching the intersection and deciding that the light will remain green long enough to get across the intersection. After making the decision to proceed through the intersection, the subject driver observes a left-turning vehicle decelerating in the lane of travel, which requires the driver to change lanes before crossing through the intersection.

This scenario is divided into four segments (Approach, Prepare for Lane Change, Execute Lane Change, and Intersection Entry). The primary reason for parsing this scenario into these segments is that each segment has a different overall driving goal (table 41). Unlike other scenarios, the speed does not vary greatly between segments.

Table 41. Scenario 4–Straight on Green Light driving objectives and speed characteristics for each scenario segment as a basis for the scenario partitioning.

Because the light remains green throughout this scenario, the dilemma-zone or red-light-running crash issues described in Scenario 3 have minimal relevance for identifying potential areas of difficulty. On the other hand, the lane-change element of this scenario provides a potential source of insight into the factors that might lead to crashes in this scenario. Lane change/merge-related crashes accounted for 3.7 percent of all crashes according to 1991 General Estimates System (GES) data, and 22.6 percent of those occurred at intersections or were intersection related.⁽²¹⁾ Most of these crashes involved little or no longitudinal gap and a small speed differential between the subject and the POV. Also, crashes were roughly equally distributed among collisions with the front (31.7 percent), middle (26 percent), and rear of the POV (35 percent). The primary causal factors were “looked but did not see” (61.2 percent) and “misjudged” gap/velocity (29.9 percent). This finding suggests that the tasks associated with determining whether it is safe to change lanes in the Prepare for Lane Change segment would be key performance bottlenecks in this scenario, especially with the inherent time limitations created by the approaching intersection. Figure 29 lists Scenario 4 details and shows the scenario diagram.

Several assumptions were made about the situational aspects of the scenario. The justifications for these are summarized in figure 29 and more fully described in table 42.

Table 42. Scenario 4–Straight on Green Light assumptions and corresponding justifications.

Scenario 4 Timeline

An approximate timeline showing the key temporal milestones for Scenario 4 was calculated based on vehicle kinematics (figure 30). These milestones were used to make judgments about the pacing of tasks within segments in addition to providing a basis for the overall sequencing of certain tasks.

Task Analysis Table

The results of the task analysis organized by scenario segment are shown in the task analysis table (table 43). The task analysis results are duplicated for individual segments in the segment analyses tables in the next sections, which more fully discuss the organization and content of the tasks and information processing subtasks.

Table 43. Scenario 4–Straight on Green Light task analysis table.

Segment Analysis

Scenario 4, Segment 1, Approach

The Approach segment involves the subject vehicle traveling at full speed until the subject vehicle driver observes that the lead vehicle is signaling to turn (which prompts the decision to consider changing lanes-next segment). The tasks, information processing subtasks, and workload estimates associated with this segment are shown in table 44. The scenario diagram, relative timing of tasks, and potential contributions to information processing bottlenecks and mitigating factors are shown in figure 31 and table 45.

Some assumptions about the task analysis and workload estimation warrant discussion. One is the assumption that the pedestrian signal is visible to the driver during the approach, and it displays the walk signal indicating that the light will remain green in the near future. Observing the pedestrian signal was mentioned as a common strategy in intersection approaches by drivers of all ages in the focus group component of this project.⁽²²⁾ This assumption consequently simplifies task 4.1.7 (determine if the light is about to change) and provides some variety from the other scenarios. It is also the most likely situation because the light is assumed to be early in the green phase. Another assumption is that the lead vehicle does not start signaling the turn until after the subject vehicle has completed task 4.1.6, which occurs because this event is the boundary between this segment and the next.

Task Pacing and Timing - Task 4.1.1 (lane maintenance) is forced-paced because it is part of the ongoing task of driving. Task 4.1.7 (observe that lead vehicle is signaling a left turn) is forced-paced because the driver has a limited amount of time to notice the turning vehicle and decide to change lanes because the intersection is rapidly approaching.

Regarding the task ordering, tasks 4.1.3 through 4.1.6 are sequential and can essentially be performed in any order. The ordering chosen in this segment followed a logical sequence and was consistent with the ordering of similar tasks in other scenario Approach segments. Task 4.1.7 was simply assumed to occur last because it was the transition point to the next segment.

Scenario 4, Segment 2, Prepare for Lane Change

The Prepare for Lane Change segment spans the time from when the subject driver observes the lead vehicle signaling to turn to when the driver decides that it is legal and safe to make a lane change. The tasks, information processing subtasks, and workload estimates associated with this segment are shown in table 46. The scenario diagram, relative timing of tasks, and potential contributions to information processing bottlenecks and mitigating factors are shown in figure 32 and table 47.

Table 46. Scenario 4–Straight on Green Light Prepare for Lane Change segment of tasks and information processing subtasks.

* Difficulty is increased by a value of 1 because of degraded information.

Several points about the task analysis and workload estimation warrant discussion. The first is that task 4.2.4 (determine whether there is a sufficient gap in right lane) was added to this segment because it is first necessary to determine whether there is room to change lanes and if a speed adjustment is necessary before checking for rear-approaching traffic.⁽²⁴⁾ It is also noteworthy that no separate task is defined for the actual decision about whether to change lanes. This decision process is the culmination of tasks 4.2.3 through 4.2.6, and the assumption is that the decision is an implicit result of resolving these tasks. Finally, the workload value of the task 4.2.5 (check rearview mirror for rear-approaching traffic) cognitive subtask was incremented because the driver must deal with degraded information from the rearview mirror.

Task Pacing and Timing - Task 4.2.1 (lane maintenance) is forced-paced because it is part of the ongoing task of driving. Tasks 4.2.3 through 4.2.6 are forced-paced because they must be completed before initiating the lane change, which is itself constrained by the approaching intersection.

In the task ordering, tasks 4.2.3 through 4.2.6 are shown as overlapping sequential tasks that follow the sequence described in the McKnight and Adams task analysis.⁽⁹⁾ The reason for presenting them in this way is that these tasks all involve visual information acquisition (which can only be done in sequence) that requires precise yet rapid deployment of visual gaze and attention to distributed locations throughout the visual scene. Consequently, there is likely to be some interference between sequential tasks; however, because more time is available than in other time-limited segments with task interference (e.g., Decision to Proceed in Scenario 3), the overlap was limited to adjacent tasks, rather than making them all simultaneous.

Scenario 4, Segment 3, Execute Lane Change

The Execute Lane Change segment spans the time from when the subject vehicle initiates the lane change until the driver is established in the new lane. The tasks, information processing subtasks, and workload estimates associated with this segment are shown in table 48. The scenario diagram, relative timing of tasks, and potential contributions to information processing bottlenecks and mitigating factors are shown in figure 33 and table 49.

* Difficulty in this subtask is increased by a value of 1 because of degraded information.

Several aspects of the task analysis and workload estimation warrant discussion. The first is that task 4.3.3 (activate turn signal) was included in this segment because this task should be performed after determining that it is legal and safe to change lanes. Although this task could have been included in the previous segment, it was included here largely because research indicates that in practice, many drivers do not activate the turn signal until the lane change is underway.⁽²⁵⁾ Also, the turn signal activation workload demands were assigned the minimum value of 1 because this is a learned automatic activity. Another point is that the maintain safe lane position task was not included in this segment because the subject vehicle is changing lanes during this segment.

One assumption in this segment is that tasks 4.3.5 and 4.3.6 (avoiding conflicts with lead and following vehicles in the right lane) do not conflict (e.g., that a speed change required to avoid the lead vehicle will not cause a conflict with the following vehicle) because it is likely that the driver would have decided earlier that there was an insufficient gap to change lanes. Finally, the estimated workload value for the task 4.3.5 cognitive subtask was incremented by a value of 1, because determining whether the relative speed and distance of the following vehicle are safe is more difficult to do using the degraded indirect visual information from the rearview mirror.

Task Pacing and Timing - Tasks 4.3.2 through 4.3.5 are forced-paced because they must be completed before the subject vehicle gets too close to the intersection.

In the task ordering, task 4.3.2 (activate the turn signal) initiates the segment, whereas tasks 4.3.3 and 4.3.4 are ongoing both before and during the actual lane-change maneuver (task 4.3.5).

Scenario 4, Segment 4, Intersection Entry

The Intersection Entry segment spans the time from when the driver has completed the lane change and is established in the new lane to when the subject vehicle crosses to the other side of the intersection. The tasks, information processing subtasks, and workload estimates associated with this segment are shown in table 50. The scenario diagram, relative timing of tasks, and potential contributions to information processing bottlenecks and mitigating factors are shown in figure 34 and table 51.

* Difficulty in this subtask is increased by a value of 1 because of degraded information.

Task 4.4.4 (check that there will be no conflict with following vehicle) is included in this segment as a quick doublecheck for potential conflict with the following vehicle. This action is a check for the possibility that the subject vehicle has misjudged the available space, or that the following vehicle was distracted and adopted an unsafe trajectory. In the McKnight and Adams task analysis,⁽⁹⁾ the corresponding task is designed to check to see if the following vehicle is trying to overtake or pass. This situation is unlikely to apply this close to the intersection; thus, the purpose of task 4.4.4 was just to doublecheck for conflict.

Also, tasks 4.4.3 (maintain safe distance from lead vehicle traveling at constant speed) and 4.4.4 cognitive subtasks were treated as involving the evaluation of a single dimension (workload estimate = 4), rather than as the evaluation of multiple dimensions (workload estimate = 6) based on relative speed and distance. The reason for this type of treatment is that evidence suggests that drivers may evaluate time-to-arrival as a single integrated variable (tau) rather than as separate speed and distance components.⁽¹¹⁾ In addition, the estimated workload value for the task 4.4.4 cognitive subtask was incremented by a value of 1, because determining the trajectory of the following vehicle is more difficult to do using the degraded indirect visual information from the rearview mirror.

Task Pacing and Timing - Task 4.4.1 (lane maintenance) is forced-paced because it is part of the ongoing task of driving. Tasks 4.4.5 (check for red-light-running traffic) and 4.4.6 (check for oncoming turning vehicles) are forced-paced because they have to be completed as the subject vehicle is rapidly approaching the intersection.

In the task ordering, task 4.4.3 (maintain safe distance from following vehicle) occurs first because this task should be performed soon after arriving in the new lane because its purpose is to doublecheck for potential conflicts. Also, tasks 4.4.5 and 4.4.6 are ordered based on which one is encountered first, and the other tasks are concurrent.

Scenario-Wide Analysis

To help identify potential information processing bottlenecks in this scenario, workload estimates from all segments were combined into a single scenario-wide workload profile that provides a general indication of where the areas of high workload demands are likely to be.

Figure 35 shows the summed workload estimates (separately for each information processing subtask) in each segment interval for the entire scenario. Also, the intervals in which key tasks are forced-paced are shaded in orange. As indicated by figure 35, the workload levels peak at moderate to high levels for the perceptual elements during the Prepare for Lane Change segment and remain at moderate levels beyond that. The cognitive elements peak at moderate to high levels at the Execute Lane Change segment and remain at moderate levels during the Prepare for Lane Change and Intersection Entry segments.

Figure 36, which displays the average workload estimate of all the tasks in play during a particular segment interval, shows that there is a peak at high levels for cognitive elements during the Execute Lane Change segment. This high average workload level associated with the underlying tasks is likely responsible for the corresponding peak in the combined workload profile for cognitive tasks (figure 35). Similarly, average workload levels are relatively high for the perceptual elements in all segments, but peak in the Prepare for Lane Change segment, which suggests that the corresponding peak in the combined workload profile (figure 35) arises from high workload levels of the tasks, rather than solely from a high number of overlapping tasks.

Intervals containing nonroutine forced-paced tasks are shaded in orange. This graph shows the overall level of workload associated with a segment.

Figure 35. Scenario 4–Straight on Green Light total estimated workload ratings for all tasks in each scenario segment.

Intervals containing nonroutine forced-paced tasks are shaded in orange. This graph generally represents the overall level of difficulty associated with the tasks in a segment.

Figure 36. Scenario 4–Straight on Green Light average estimated workload ratings per task for each scenario segment.

Information Processing Bottlenecks

Information about the combined and average workload ratings, pacing of key tasks, and nature of bottlenecks for each segment is shown in table 52. Only information that represents potential problems is listed; blank cells indicate that no substantive issues occurred for a particular segment or cell. Following the table is a list of key information processing bottlenecks identified in each of the segments.

Prepare for Lane Change nature of bottleneck: High time pressure:

• Several tasks involving moderate to high perceptual and cognitive workload focused on obtaining information from the environment are needed to decide whether or not a lane change can be made. These tasks must be performed in rapid succession under high time pressure because the time available to make the decision is limited by the approaching intersection.

Execute Lane Change nature of bottleneck: High time pressure:

• There are concurrent and ongoing perceptual and cognitive tasks with moderate to high workload conducted under time pressure because the subject vehicle must safely get into the new lane with sufficient time to perform the Intersection Entry tasks. These effects are mitigated somewhat because most of these tasks are routine, automatic activities.

Intersection Entry nature of bottleneck: Distributed information:

• There are perceptual and cognitive tasks with moderate to high workload that involve information distributed throughout the visual scene with time pressure for some of the safety-related tasks. These effects are partially diminished because some of these tasks are routine, automatic activities.

Previous | Table of Contents | Next