Cooperative Adaptive Cruise Control Human Factors Study: Experiment 3—The Role of Automated Braking and Auditory Alert in Collision Avoidance Response

CHAPTER 2. METHODS

PARTICIPANTS

A total of 112 participants completed the study, 28 in each of 4 groups. Participants in experiment 1 were excluded from participation in experiment 3. To roughly balance the groups on participant age, half the recruits in each experimental group were under the age of 47. Table 2 provides the mean, minimum, and maximum ages of the participants in each of the experimental groups. Each group shown in table 2 had 14 males and 14 females.

SIMULATOR

The experiment was conducted in the Federal Highway Administration (FHWA) Highway Driving Simulator. The simulator consisted of a compact sedan mounted on a 6‑degree of freedom motion base placed within a cylindrical projection screen with a radius of 8.9 ft (2.7 m). Three projectors were used to provide a 200- by 40-degree (horizontal by vertical) field of view. Each projector provided a nominal resolution of 4,096 by 2,400 pixels. The motion base was tuned to optimize realistic perceptions of longitudinal acceleration and deceleration and minimize false lateral acceleration cues. Prior to the experiment, a panel of sixdrivers rated the acceptability of the motion cues to be 6, where 1 represented “totally unacceptable” and 7 represented “very acceptable.” The simulator’s steering was also tuned prior to the experiment so that constant steering corrections were not required to maintain a straight path.

The simulated vehicle was equipped with a hands-free intercom system that enabled communications between the participant and a researcher who ran the experiment from a control room. The researcher in the control room could also view the face video from the eye-tracking system and thereby monitor the participant’s well-being.

SCENARIO

Participants drove in a dedicated center lane on a simulated eight-lane interstate highway (fourlanes in each direction). The center lane in the participants’ direction of travel was separated from the other lanes by F-type barriers, which are shown in figure 1. Entrance to the center dedicated lane was accessed from the left side of the roadway from a ramp with a ramp meter. The simulation began with the participant’s vehicle in the third position within a platoon of four vehicles. When the ramp meter turned green, the platoon accelerated, merged into the dedicated lane, and cruised at 70 mi/h (113 km/h). Vehicles in CACC mode were set to maintain a 1.1-s gap. For the first 5.8 mi (9.4 km) or 5 min, the platoon proceeded as formed. At 5.8 mi (9.3 km), a CACC vehicle merged into the platoon from the left in front of the participant. The merge was from a ramp identical to the initial ramp. The initial gap between the participant and the merging vehicle was about 0.5 s or 51 ft (15.5 m). The CACC or ACC systems immediately responded by decelerating to restore a 1.1-s gap.

Figure 1. Screenshot. A typical section of the simulated roadway.

There were left access and exit ramps for the dedicated lane every 2 mi (3 km). At the 11th access ramp, a vehicle traveled rapidly down the ramp and entered the dedicated lane ahead of the platoon’s lead vehicle and overturned. The overturn event was occluded from view by the three platoon members ahead of the participant. As the overturning began, the platoon’s lead vehicle began a constant deceleration of 32 ft/s² (9.8 m/s²⁾. After a 0.1-s delay, all the remaining CACC vehicles in the platoon began a constant deceleration of 0.4 g. All the CACC vehicles except the participant’s vehicle began 1-g deceleration 1 s after the 0.4-g deceleration began. The participant vehicle in the ACC condition began decelerating at 0.2 g 0.4 s after the platoon lead vehicle began hard braking (0.3 s after the vehicle directly ahead began its 0.4-g deceleration).

The CACC groups that received the auditory warning were presented a 1,000-Hz warning tone of four beeps with each beep duration lasting about 140 ms and separated by about 22 ms of near silence. The alarm was triggered at the same time as the automated braking. The CACC groups that had automated braking enabled braked at the same time and with the same 0.4-g rate of deceleration as the other CACC vehicles that responded to the lead vehicle braking event.

DEPENDENT MEASURES

The dependent measures were as follows:

• Participant crashes (collision yes/no) at the final crash event.
  
  
• Reaction time to the final event (onset of brake pedal depression).
  
  
• Minimum time to collision (TTC).

A crash was recorded if the participant’s gap to the preceding vehicle decreased to zero and the lateral position with respect to the preceding vehicle was less than the width of the design vehicle. That is, if the participant vehicle either braked sufficiently to avoid contact with the vehicle ahead or successfully swerved to avoid contact, then there was no crash.

Reaction time was calculated for either a braking response or a steering response, whichever came first after the onset of the platoon leader braking. A braking response was scored if the brake pedal position exceeded 0.02 on a scale from 0 to 100. A steering wheel reaction time was recorded if the steering wheel torque exceeded 1.125 lbf (5 N).

Minimum TTC is the adjusted minimum TTC described by Brown and also described in the experiment 1 report.^(3,4) This measure has the advantage of being interpretable even if a crash occurs; positive values indicate severity of the near crash event, whereas negative values indicate the severity of crashes. In all cases, smaller values are more severe. More accurately, positive values indicate how much extra time the participant had available to react, and negative values indicate how much shorter the reaction time needed to be to avoid a collision. If the deceleration of the following vehicle is less than the deceleration of the lead vehicle, which is still moving, then adjusted TTC goes to negative infinity and precludes its use in estimating group means.