A Precrash Scenario Analysis Comparing Safety Performance across Autonomous Vehicle Driving Modes

Table 2

Descriptive statistics of California AV crash dataset (sample size = 484).

Variable	Frequency	Percentage (%)
Someone injured
Yes	73	15.08
No	411	84.92
Vehicle damage
None	38	7.85
Minor	368	76.03
Moderate	71	14.67
Major	7	1.45
Vehicle manufacturer
Waymo LLC	193	39.88
Cruise LLC	176	36.36
Other	115	23.76
Production year of the AV
2016	71	14.67
2017	117	24.17
2018	7	1.45
2019	81	16.74
2020	71	14.67
2021	124	25.62
2022	13	2.69
Number of vehicles involved
1	77	15.91
2	402	83.06
3	5	1.03
Movement preceding crash
Stopped	188	38.84
Proceeding straight	199	41.12
Making left turn	41	8.47
Making right turn	35	7.23
Backing	21	4.34
Type of crash
Head-on	87	17.98
Rear-end	230	47.52
Left crash	85	17.56
Right crash	82	16.94
AV driving mode
Autonomous driving	234	48.35
Conventional driving	168	34.71
Driving takeover	82	16.94
Facility type
Intersection	238	49.17
Road segment	231	47.73
Parking lot	15	3.10
Time of the crash
Morning peak 7:00–9:00	47	9.71
Daytime 9:00–17:00	273	56.40
Evening peak 17:00–19:00	42	8.68
Night 19:00–7:00	122	25.21
Weather
Clear	432	89.26
Rainy	14	2.89
Cloudy	33	6.82
Fog/visibility	5	1.03
Light
Daylight	357	73.76
Dark with street lights	122	25.21
Dark without streetlights	5	1.03