CANsee Automobile Intrusion Detection System to be Released at HITB Security Conference Next Month
Amsterdam – 25th April 2016: Hack In The Box, organizer of an international series of deeply technical security conferences, returns once again to Amsterdam with an even more feature packed agenda ranging from brand new vulnerabilities in the Android ecosystem, 9 new sandbox escapes affecting cloud and virtualization providers to a talk on detecting abnormal behaviours in automobile systems using machine learning.
Our cars and automobiles have become more than just mere modes of transportation. They are now virtually computers on wheels, equipped with advanced systems enabling satellite navigation, enhanced in-car entertainment and access to the Internet - all connected through the vehicle’s controller area network or CAN bus.
Research into the security of the CAN bus is not new with Charlie Miller and Chris Valasek most famously hacking a Ford Escape and a Toyota Prius in 2013 via the vehicle’s onboard diagnostic port. In 2015, they were able to go one step further by taking control of a vehicle remotely over the internet. This included taking over the steering wheel, stopping the car, and even worse, disabling the brakes. At the time of their research, it was estimated around 471,000 vehicles were vulnerable to such attacks.
With so much ongoing research focused on breaking into automobiles, very little is currently discussed about protecting vehicles from outside attacks. One security researcher however has set out to develop a system to help detect abnormal behaviour in an automobile’s CAN bus. Mr Jun Li, a researcher with China’s Qihoo360 research team, has created a machine learning based intrusion detection system (IDS) for automobiles he claims is capable of detecting abnormal traffic. Mr Li, as part of Qihoo360’s Unicorn Team, one of the largest Chinese security companies, was also previously involved in hacking Tesla Motors’ Model S in 2015 at the Syscan security conference.
At HITBSecConf in Amsterdam however, Mr. Li will not only demonstrate the inner workings of this new IDS system he calls CANsee, but details on the project, including software used, schematics and bill of materials will be open-sourced and released to the public allowing for anyone to build a similar device themselves.
Slides and supporting material from the presentation will be released immediately after the talk and linked to the presentation page here: http://conference.hitb.org/hitbsecconf2016ams/sessions/cansee-an-automobile-intrusion-detection-system/
