Tevora’s Truman Kain, a Senior Information Security Analyst, has created Surveillance Detection Scout – Your Lookout on Autopilot, also known simply as SDS. Scout is a hardware and software stack currently supporting Tesla models S, 3 and X, meant to alert you in real time if you’re being followed while driving. This is achieved by running license plate recognition on 3 camera streams simultaneously. Scout was created to help ease the burden of mentally taxing countersurveillance measures including surveillance detection routes, however the framework has offensive applications as well.
As a result of Scout remaining vigilant and running familiar face detection while parked, a user is able to run recon queries on all previously captured data to gain an overview of observed patterns of life, during a physical penetration test or red team engagement for example. In order to mitigate the risk of user data being publicly or inadvertently exposed, Kain has purposely left all cloud storage integrations out of the SDS software stack, opting to keep computer vision and storage tasks local to the user’s chosen hardware device.
Scout is designed to run on an Nvidia Jetson Nano or Nvidia Jetson Xavier Mini Computer which is modified to act as a USB Flash Drive. This allows local inference to begin immediately on all dashcam video provided by the Tesla vehicle in which Scout is located and alert the user of any findings via mobile push notification and In-Car Browser pop-up. The price options for SDS hardware range from $100-700 USD.
The Surveillance Detection Scout project was inspired by Wagging the Tail: Covert Passive Surveillance, a talk that Kain attended last year at DEFCON 26. He spoke with one of the presenters during the conference about the idea of potentially building an automated surveillance detection solution using a mobile device, but ultimately decided that this wasn’t plausible. In February of this year, shortly after the release of Tesla’s Sentry Mode, Kain realized that a seamless surveillance detection solution was now possible using his Model 3.
Sentry Mode effectively allows the Tesla vehicle on which it is enabled to provide uninterrupted power to devices plugged in via USB or car charger. As a result, Scout is able to run 24/7, assuming that the vehicle is regularly charged. Sentry Mode, along with the built-in dashcams, make Scout “a natural fit for the Tesla,” said Kain. By assuming the role of a USB flash drive, Surveillance Detection Scout proves that even a seemingly basic feature like the dash camera can be utilized for a much less consumer purpose: aiding in surveillance and counter-surveillance.
Possibly one of the most surprising details of this project is that only freely available computer vision frameworks for face and license plate recognition were utilized. “I’m not doing any cutting-edge AI, I’m just applying what’s already freely available, off the shelf,” said Surveillance Detection Scout creator and Tevora Senior Information Security Analyst, Truman Kain.
Surveillance Detection Scout – Your Lookout on Autopilot was released on Friday August 9, 2019 at DefCon 27.
Truman Kain is a Senior Information Security Analyst at Tevora, where he specializes in Social Engineering penetration tests. Kain has a background in design, marketing, and mobile development, which he utilized to develop Dragnet, an intuitive phishing framework. This year, he has combined his knowledge of machine learning and design experience to create Surveillance Detection Scout.
Founded in 2003, Tevora is a specialized management consultancy focused on cybersecurity, risk, and compliance services. Based in Irvine, CA, our experienced consultants are devoted to supporting the CISO in protecting their organization’s digital assets. We make it our responsibility to ensure the CISO has the tools and guidance they need to build their departments, so they can prevent and respond to daily threats.
Skylor Phillips, PR Coordinator