Carnegie Mellon University researchers have developed a technology that detects three dimensional shapes and movements of human bodies in a room, using only WiFi routers. By analysing radio waves reflected off bodies, their system reconstructs human bodies without cameras or LiDAR.
This technology leverages artificial intelligence and existing WiFi infrastructure to provide an alternative to cameras, overcoming their limitations. This technology and its implementation is meant to address challenges in privacy, cost and environmental adaptability. It is redefining how we view privacy, surveillance, and monitoring and brings concerns of invasive practices being implemented.
How WiFi-Based Human Detection Works
The technology leverages Channel State Information (CSI), which is data capturing how Wifi signals distort as they interact with objects. The system uses three routers and receivers to track phase and amplitude changes across signals. These signals are measured and measurements are then organized into a format that shows how movement distorts the radio waves. This creates a detailed picture of these bodies’ movements that affect the signal.
Researchers adapted DensePose technology for this system of mapping. Initially developed by London and Facebook’s AI team, DensePose maps human bodies in photos. Carnegie Mellon’s team integrated it with a deep neural network to process this data. The neural network then links WiFi signals in phase and amplitude. This translates WiFi signals into precise coordinates on human bodies. It filters static objects, focusing on moving bodies. Data becomes radar-like images, reconstructing poses in 3D. Unlike cameras, this method is cost-effective and works without light.
Advantages Over Traditional Sensing methods
The use of WiFi-based sensing mitigates the limitations of traditional methods. While RGB cameras struggle with low-light and LiDAR systems are expensive, routers are an inexpensive option, being 90% cheaper than LiDAR. Unlike cameras, routers function regardless of lighting conditions. Signals can also penetrate walls, expanding its reach. This enables reliable monitoring in cluttered environments.
Researchers assert that “WiFi signals project individuals’ privacy while offering functionality” since the system detects geometric distortions rather than capturing visual details. Traditional cameras are absent and so it avoids capturing identifiable images, preserving anonymity.
The researchers position it as ideal especially for sensitive settings such as bathrooms or elderly homes.
Read More: Is It Better to Place Your Phone Screen Up or Down on the Table?
Potential Applications and Their Dangers
Researchers envision uses in health and elderly care, as well as in security applications. The system could track movements and detect falls among seniors instantly, triggering alerts without intrusive camera surveillance. Hospitals and healthcare facilities can monitor patient mobility during recovery.
Home security is another potential application. It might identify “suspicious behaviors” in homes. However, the research vaguely defines what is considered “suspicious behavior”. These applications highlight versatility, yet this ambiguity in defining behaviors raises ethical alarms. Critics note that with clear thresholds, such systems could flag benign behaviors like nighttime movements as suspicious. Companies like Amazon have already faced scrutiny for Ring cameras, which could exploit WiFi tracking for passive home surveillance.
Applications and Dangers of Unwanted Surveillance
While promising, this method still has its limitations. Training data came from controlled environments with limited diversity, risking biased performance in real-world settings. The neural network relies on precise CSI calibration, where devices like microwaves or phones can interfere with the CSI.
Researchers claim this system protects privacy by avoiding visual recordings. So although no images are stored, the system catalogues data on movement patterns which could reveal individuals personal habits. Critics warn that this technology could breed new forms of surveillance. AI models can also introduce biases and inaccuracies.
Widespread WiFi-based detection could amplify corporate and governmental overreach. The ambiguity around “suspicious behaviors” creates room for abuse from authorities. This lack of clarity as to what is deemed “suspicious” demands careful consideration before implementation. However, researchers acknowledge the necessity of encryption and anonymization, proposing to implement these tactics for future projects.
Future Trajectory and Scrutiny of Surveillance
This concept of WiFi-based tracking builds on a decade of research. In 2013, MIT used cell signals to detect humans through walls. Five years later, another MIT team used WiFi to map movements as stick figures in adjacent rooms. Carnegie Mellon’s work advances this by achieving greater detail and precision, rivaling camera-based pose estimation.
Future iterations could integrate with smart home systems, offering energy savings by adjusting thermostats based on bodies in the room. Public spaces might deploy the technology for crowd analytics without facial recognition. However, regulatory frameworks lag behind. Clear guidelines are needed to prevent misuse by landlords, employers, or law enforcement.
Retailers could analyze customer behavior without cameras. Event organizers might track crowd movements. Emergency responders could locate trapped individuals. However, realizing these possibilities requires ethical balance. We must consider the potential for misuse before widespread deployment.
Weighing Innovation Against Privacy with WiFi
Carnegie Mellon University’s WiFi-based human detection technology offers a promising alternative to traditional cameras by using radio signals to map human movements in 3D. Unlike cameras, this system operates invisibly, avoiding visual surveillance while overcoming challenges like poor lighting.
However, its passive monitoring capabilities raise concerns about privacy, as households could unknowingly surrender detailed movement data to third parties. Researchers emphasize the need for ethical safeguards, including transparent use cases, strict data ownership rules, and opt-out mechanisms, to prevent misuse. Without these protections, the widespread adoption of WiFi tracking risks turning everyday routers into covert surveillance tools. Balancing innovation with privacy remains critical as this technology advances.
Read More: Should You Turn Off Your Smartphone’s Wifi and Bluetooth At Night?
