COLLEGE PARK, Md. – Robots can make daily chores much simpler than cleaning the house by hand. Unfortunately, a new study reveals that they can also make things much easier for hackers to steal personal information. An international team has discovered that robot vacuum cleaners can actually be reprogrammed remotely to record sound waves, even though they don’t have a microphone.
Researcher Nirupam Roy of the University of Maryland says the study has successfully gathered information using the laser navigation system of a household vacuum robot. The team found that the robot’s laser signals could be converted into sound waves. With the help of signal processing software and machine learning systems, computers can accurately select speech and music patterns from television programs that the vacuum eavesdrops on.
“We welcome these devices into our homes and don̵
The study found that these devices rely on light recognition and range (Lidar) to help them rotate around a room. The vacuum illuminates its laser in a room and senses the reflections as the beam bounces off walls and objects. The machine can then outline where to go and when to turn for effective house cleaning.
While this technology is great for sensing the robot’s direction, security experts have already alerted where this information is stored. The study reveals that mapping information is often stored in the cloud. This means that a breach of privacy may reveal information about the size of the consumer’s house. Roy suggests that third parties may gain an unauthorized view of the income and lifestyle of unsuspecting vacuum users.
The researchers set out to find out if Lidar could also pose a threat as a real-time recorder. The study explains that sound waves cause objects to vibrate. This actually causes small variations in the light reflected from them. Since the 1940s, laser microphones have been able to convert these variations in light back into a sound wave. These devices, often used in espionage operations, rely on the laser beam reflecting on a smooth surface like a window.
Unlike professional eavesdropping equipment, robotic vacuums illuminate their lasers on objects that often have an irregular shape or density. Roy’s team wasn’t sure if the distracted signal would provide enough information to restore the original sound wave.
Turn a robot vacuum cleaner into a spy bot
To test whether the vacuum could actually record sounds, the researchers first penetrated the robot’s software. This revealed that it is possible for someone to control the position of the laser beam and receive this data using a laptop with Wi-Fi.
The team then tried to recover sound patterns from two different audio sources. One involved a human voice reciting numbers through a computer column. The other comes from a TV that broadcasts various programs while the machine is being cleaned.
The study then captures Lidar signals when they bounce off several nearby objects, including a trash can, a cardboard box, a cardboard box and a shopping bag. Researchers have passed the signals through their computer-based learning algorithms, which can distinguish the difference between the human voice and music.
The results reveal that the team’s computer program, called LidarPhone, can identify 90 percent of the numbers spoken in the room. The program can also successfully select 90 percent of TV shows from a one-minute recording.
“This kind of threat can be more important than ever when you think we all order food over the phone and have appointments on the computer and often talk about our credit card or bank information,” warns Roy.
“But what worries me even more is that it can reveal much more personal information. This kind of information can tell you about my lifestyle, how many hours I work and other things I do. And what we watch on television can reveal our political orientations. This is crucial for someone who may want to manipulate political elections or send very specific messages to me. “
It’s not just the vacuum cleaner that can spy on you
The team warns that robotic vacuum cleaners are just one possible source for Lidar to spy on the market today. Other devices that use an infrared sensor, such as smartphones, can potentially collect information without your knowledge. Many of these devices use their sensors to detect faces or detect movement.
“I believe this is significant work that will make manufacturers aware of these opportunities and will trigger the security and privacy community to come up with solutions to prevent such attacks,” Roy said.
The study was presented at the conference of the Association of Computing Machines for Embedded Network Sensor Systems (SenSys 2020).