Page 122 - VERITAS Vol.2 Issue 2

Society of Automotive Engineers.

F. In 2016, Apple revealed for the first time that it is developing auton-

omous automobiles. In California, Apple boasts the third-largest fleet of

self-driving test vehicles.
G. By 2030, Kia, a division of Hyundai Motors, wants to market a com-

pletely autonomous vehicle.

FORENSIC CHALLENGES

Fig 2. Image shows the various possible entities that could be liable when an
accident occurs by an SAV

AV forensics framework imposes several unique challenges for forensics

investigation. The significant challenges of AV forensics are:

• Huge amount of data: An AV generates a huge amount of data each
day. In each second, radar and ultrasonic sensors generate 10-100 KB

(Kilobyte), GPS (Global Positioning System) generates 50 KB, cam-

eras generate 20-40 MB (Megabyte), and LiDAR generates 10-70 MB

data. On average, an AV generates 4000 GB (Gigabyte) of data each

day. Such massive amounts of sensor data impose challenges in AV

forensics.

• Log accessibility: Proper access control of information and logs col-
lected from AV is essential. The developers of AD (Autonomous Driv-

ing) systems and the investigators should be able to access the low-lev-

el information. However, any third-party malicious entity should not

113 Veritas Volume: 2, Issue: 2