Going on vacation can be synonymous with fear of burglary. In order to provide some peace of mind while keeping an eye on the fireplace, surveillance cameras are there, but they can also turn against the user.
Based on readings from Nest and Xiaomi surveillance devices, researchers from Queen Mary University in London and the Chinese Academy of Science demonstrate in a study that a hacker can measure the flow of information from an IP (Internet Protocol) camera without accessing the videos or other content recorded by it.
In practice, an IP camera transmits information using the internet stream (Wi-Fi most often), whether for direct or recording. Most of the time, this data goes through the manufacturer’s cloud and servers and is then redirected to the user (on their smartphone or tablet).
According to the researchers, if a hacker is able to track this traffic, he can know the frequency and the time of the detected activities, as well as their types. It could also identify the moments when the camera is deactivated or activated (and therefore visualize user habits). Finally, depending on the amount of data collected, the cybercrusher would be able to predict the moments when the activity around the camera is less and, potentially, when the place where the camera is positioned is unoccupied.
A real problem therefore for owners who, by using their devices, sketch their habits to malicious people. Even if this is alarming, doctor Gareth Tyson, one of the authors of the study, wants to be reassuring by explaining to CNN that to achieve this type of hack, you need an advanced level of knowledge in this area.
This study therefore highlights a major concern which, if the hack protocol is “democratized”, can call into question the usefulness of surveillance cameras, purely and simply.
Gareth Tyson ends by explaining that with the results of this study, the researchers want “Find a smarter system that would allow the cameras to understand what the detection is, assess the degree of risk and then, if the event really deserves it, record the movement while alerting the user” “. The latter would therefore no longer be inundated with unnecessary notifications and this would make the surveillance system much more relevant.
Manufacturers are already trying with various features to make their products more precise. So we can note the differentiated detection of theArlo Ultra, for example, which allows you to be warned only of movements of a category (humans, animals, vehicles), or even facial recognition which makes the camera capable of distinguishing people known from intruders, like the Netatmo Smart Indoor Camera.