Jon Atwood

Automated Vehicles: Take-Over Request and System Prompt Evaluation

As automated driving technology advances, the driver’s role continues to shift from active vehicle control to passive monitoring of the automated driving system and environment. This study comprised three experiments on controlled test tracks in mixed traffic that investigated driver interactions with prototype Level 2 and Level 3 partially automated driving systems. The study investigated which human-machine interface (HMI) characteristics are most effective at issuing a Take-Over Request (TOR) during the operation of a Level 2 automated driving system, how to prompt drivers to attend to the road when distracted during the operation of a Level 2 automated driving system (and whether these prompts are effective over time), and which HMI characteristics are most effective at issuing a TOR during the operation of a Level 3 automated driving system. In addition, participants’ trust in the automated driving system they experienced was gauged through multiple Likert-type surveys and an after-experience interview.

The driver-level crash risk associated with daily cellphone use and cellphone use while driving

This study examined the overall prevalence of cellphone use, including the rates of calls and texts both per day and hourly while driving, and assessed whether or not individual crash risk was correlated with cellphone use. The study used data from the Second Strategic Highway Research Program Naturalistic Driving Study (SHRP 2 NDS), which had more than 3500 participants who provided up to three years of driving data. Of these participants, 620 provided cellphone records, 564 of which included both call and text records. The prevalence of cellphone calls and texts per day was calculated. By overlaying the cellphone records with the SHRP 2 NDS data, we also evaluated the rates of calls and texts while driving by driver demographics. Crashes for these cellphone-using participants were also identified from the SHRP 2 NDS data. Negative binomial regression models were used to determine whether the crash rate was associated with cellphone use. Participants made an average of 27.1 texts and 7.3 calls per day. They averaged 1.6 texts and 1.2 calls per hour of driving. Cellphone use varied significantly by age, especially for texting. The texting rate for drivers aged 16-19 was 59.4 per day and 2.9 per hour of driving, four times higher than the 14.3 per day and 1.0 per hour for drivers 30-64 years old. The texting rate for drivers 20-29 years old was also high at 42.4 per day and 2.6 per hour of driving. Participants experienced 243 crashes in 216,231 h of driving. It was found that those who texted more often per day or per hour of driving had higher crash rates after adjusting for age and gender effects. The severe crash rate increases 0.58% for every additional text per day and all 8.3% for every text per hour of driving; overall crash rate increases 0.41% for every additional text per day and 6.46% for every text per hour of driving. The results show that cellphone texting and calling are quite common while driving. The texting rate for young drivers is substantially higher than for middle-aged and senior drivers. This study confirmed that those who text at a higher rate are associated with a higher crash risk.