There is one article on EnCase blog, "The Car of the Future May be a Forensic Gold Mine"Looks like a discussion about car and digital forensics started on CEIC 2014 which then spread to LinkedIn groups. Very interesting topic, also very frustrating since most of the current digital forensic tools are not up for the task. It is possible to extract data from cars but only partially, and by using available general purpose tools on forensically well know car subsystems, like GPS. There are plenty of examples of car GPS systems with other subsystems analyses as well as CAN analyses. This should provide a great improvement over early investigations with cruise system error related accidents and deaths. This story requires a lot of research, even though the case is still ongoing, due to the important fact that the relevant data was not extracted from the car systems, so we there is a serious problem there.
More recent story "BMW Fixes Software Flaw that Affected 2.2 Million Cars (February 2, 2015)"
published on SANS which shows the spread of the problem to almost the size of the fleet of mobile devices.
To be honest, modern car digital forensic more like scada system analyses than anything else. Even worse, the car systems are not designed to be forensically reliable or even computationally safe. Car systems are designed to be reliable as old mechanical control systems in cars were before. Electronics, communications and interacting electronic/computing systems makes this situation even worse.
I would recommend that anyone dealing with car forensics or security should go to Nancy Leveson's page and read few papers.
