Smart home: spies in the apartment

Smart home: spies in the apartment

For the end customer, a vacuum cleaner robot is simply a piece of electronics that promises more convenience. For security researchers and hackers, a vacuum robot is a highly interesting target. Researchers have repeatedly shown in recent years that this danger is not hypothetical. In our naïve enthusiasm for technology, have we plastered our apartments with spy tools? The answer is complicated, reports MIT Technology Review in its latest issue – but not necessarily reassuring.

Advertisement

Probably the most spectacular hack so far: In 2021, a research team from Singapore was able to show that a lidar scanner built into the vacuum robot can also be used as a listening microphone. In 2021, Sriram Sami from the University of Singapore and his colleagues replaced the firmware of a Xiaomi vacuum cleaner and tapped into the lidar sensor data to eavesdrop on users.

It sounds like James Bond, but in reality it wasn’t quite as spectacular: lidar sensors emit an infrared laser beam and measure how long it takes for the reflected beam to come back. The result is a three-dimensional “point cloud” of the environment. If an object vibrates, for example a window pane, because sound waves hit it, these sound waves can theoretically be reconstructed from the lidar data. In their experiment, however, the researchers had to apply a classifier to the data and were only able to distill the gender of the speaker, numbers and some music fragments from the data.

Security level of IoT devices

Since then, no other articles have appeared that report on similarly spectacular hacks. Have manufacturers learned from this? Yes and no, says Prof. Ahmad-Reza Sadeghi, head of the System Security Lab at the TU Darmstadt, who has been researching security in the Internet of Things (IoT) with his team for many years. In 2019, a graduate student of Sadeghi himself reported the successful hack of a then-popular vacuum cleaner robot.

In general, the security level of the devices has improved, but there is still a lot to do. “Many gaps are closed without the public noticing anything about it,” says Sadeghi, and he adds: “It’s more the small companies with particularly innovative products that sometimes neglect security.” Overall, however, there would be more material and much more experience in security issues. In general, the big providers are getting better and better at it. But this development is also double-edged: “The big companies are getting more and more data power. And we’ve seen what can happen when a company like Facebook controls all our data.”

In fact, it’s not always hacks that turn smart devices into surveillance machines. In the fall of 2020, gigworkers in Venezuela posted a series of photos on online forums where they shared their work. The photos showed everyday, if sometimes intimate, scenes from the household, shot from low angles. Manufacturer iRobot confirmed that these images were taken by its Roombas in 2020.

Advertisement

ChatGPT and Co. question the extent to which the classic transfer of knowledge in the classroom still makes sense if, in the future, an AI will provide almost all the knowledge in the world within seconds in the required form. How can schools react to this? The new issue of MIT Technology Review addresses this question. Highlights from the magazine:

“What we see today is just the beginning”

Anyone who fears that iRobot is too careless with data shouldn’t be reassured by the news that Amazon wants to acquire iRobot for $1.7 billion. The approval of the US supervisory authority Federal Trade Commission (FTC) is still pending, and the takeover is also being examined by antitrust law in Europe. Either way, the appetite for data will only increase in the coming years.

Robot vacuum manufacturers are already investing in more features and devices that will bring us closer to a robotic future. “What we’re seeing today is just the beginning,” says security expert Sadeghi. “In the future, smart devices will be more and more integrated directly into homes – not individual devices will then act, but the entire smart environment. With smart connectivity and AI accessing cyber-physical systems, we will have a much bigger problem than today with separate devices.”

Read MIT Technology Review here:

(wst)

To home page

#Smart #home #spies #apartment