There is currently a market for reprogramming performance aspects of engine controllers such as emission controls. This is mostly done in professional race circuits, but it is also an aspect of high-end street racing. It looks like this may eventually become a victim of increased emphasis on security as automotive networks get locked down. This is similar to Linux's problems with UEFI BIOS on PCs, where security concerns removes some degree of what a user can do with something that they own.
Will this extend to other third-party add-ons as well? Will a third-party stereo be able to get access to light levels or other information that a user might want it to? Will that be a big concern for car companies that have never liked third-party add-ons anyway?
Perhaps it's just me. I tend to get exasperated when this topic keeps piling on the various security holes as if they were all critical. Hacking via a physical connection to the OBD-II port, or hacking into the infotainment system to change the station, or hacking into the telematics system to see how fast you're going, are no more of a threat than someone listening in on your cell phone conversation. That can happen too, yet we're not making a huge deal about it.
On this topic, it is just this sort of sensationalism that turns me off. I wouldn't be surprised if others feel the same way.
On the other hand, the real threats are to the critical systems, as I've suggested in the past, i.e. brakes, steering, and lastly throttle, and in particular with a remote wireless link. Not physical connection to OBD-II, even though these articles keep insisting that should count too. As many have said, an inside the car physical sabotage can take such a gymongous variety of types that it strains credulity to give an electronic attack any emphasis. It just sounds like a desire to sensationlize. So let's focus.
It seems that at least one car model allows disabling of the brakes via a remote wireless link. To me, that's inexcusable. The steering vulnerability that was uncovered, on the other hand, appears to be manually overridable. That's good design, and why not mention that explicitly?
I'm not denying the importance of introducing more security protocols into automobile electronics as these become more interconnected and pervasive. Just asking to increase the signal to noise ratio, as it were.
I agree, goafrit. Even more interesting to me is that a lot of security issues chip companies have learned by working with other industries --computers, smartcards, mobiles -- are now becoming very relelvant and applicable to the automotive market.
Automobile security is a new fast developing market with all the mobility and infotainment. It is very refreshing that TI and co are not leaving that for firms like Bosch, TRW, and Conti to dominante as has been the case in many automotive areas. Now, they are moving fast.
Blog Doing Math in FPGAs Tom Burke 24 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...