Many burglaries remain unsolved today. Thanks to the spread of smart-home connectivity and all manner of new sensors, investigators have more tools at their disposal—at least in theory.
Modern TV detective dramas have normalised the idea that, upon arriving at a crime scene, the inspector quickly asks underlings to check various surveillance cameras. But in a smart home, there may be far more devices tracking activity: robotic vacuums, lamps with built-in motion sensors, thermostats, and more. Many of these devices transmit or store measurements—sometimes to the cloud, sometimes on a local platform such as Home Assistant. This leads to a vast array of data that could prove very useful in understanding what really happened during a break-in or other crime.
Wider Prospects in IT Security
Alongside general cyber-defence measures and the capacity to analyse safety-critical data, the concept of using smart-home devices as “witnesses” is just one example of new investigative frontiers. Various institutes and labs, and university research hubs—are collaborating to ensure the police aren’t overwhelmed by the rapid pace of consumer tech innovation. After all, with so many new devices and standards popping up, an investigator could easily overlook valuable clues hidden inside a home’s interconnected sensors.
Smart Meters as Clues
A single sensor reading every quarter of an hour doesn’t sound like much, yet even that can indicate a suspicious fluctuation in electricity usage. For instance, if a 500-watt halogen lamp is triggered by a simple motion detector in an otherwise quiet house, that extra usage might appear in the reported data. From the standpoint of an investigator, it can help pinpoint when, for example, the suspect might have entered the scene.
One of the first widely noted examples of such usage data helping solve a murder occurred in Bentonville, Arkansas, in 2015. Police discovered that a suspect, who claimed to have been asleep, had actually used significant amounts of electricity and over 500 litres of water during the night—evidence pointing to him cleaning away incriminating details on the patio. Smart-meter readings ended up contradicting his story and contributing to his conviction.
The Spectrum from Sparse to Rich Data
Smart meters alone provide a limited glimpse of a crime scene, but camera footage at the other extreme might show the entire act and the faces of those involved. Between these two lies a diverse array of home sensors: everything from motion and presence detectors to vacuum cleaners that map rooms and react to the presence of people.
Systems like the Bosch Smart Home Controller or IKEA’s Trådfri line illustrate how quickly sensor suites are expanding. Thermostats may rely on door or window sensors that detect open windows. Motion sensors can switch lights on when they detect movement, and in some cases they also feed data to thermostats or ventilation settings. Air-quality sensors, meanwhile, can gauge how many people are present, since more exhaled carbon dioxide typically signals increased occupancy. And there are dedicated security devices, such as glass-break sensors or door contacts, plus the rising popularity of voice commands for operating blinds, music, or lights.
If there’s an ideal scenario for investigators (aside from an actual camera recording the crime), it might be a smart enthusiast’s home with a system like Home Assistant logging every sensor reading. Over time, that database can show precisely which sensors triggered at what moment—indicating how many people were in which rooms and when. Add the wireless router logs that show which smartphone tried to connect at a given time, and a detective may have robust leads on the entire timeline.
Of course, for serious crimes, police already cross-check phone records of suspects. They can also request voice data from Amazon’s Alexa or other assistants, given the right court order. But even if a major warrant isn’t in play, patterns in Alexa’s communications with the cloud (e.g., the times and volume of data transmissions) could reveal details, even when the content of the messages is encrypted. If the data spiked at 2am, for instance, something unusual likely happened then. By cross-referencing that with other sensor logs, investigators start to piece together the probable chain of events.
Building Blocks of Evidence
In a previous study, Büsching’s team examined how a human body can affect the signal quality in Wi-Fi or Zigbee networks if someone stands between the transmitter and the receiver. While the tests were performed in a lab setting, the principle could apply to a real home environment: if a person’s presence measurably alters the signal compared to historical data logs, it might hint that an intruder was moving around. However, this requires careful validation outside the lab.
Not all of this research concerns only the most serious crimes. Burglaries are widespread and frequently go unsolved. Victims often lose not just valuables but their sense of security. Many of those with expensive homes or possessions have likely embraced early smart-home tech, so it stands to reason that data gleaned from these systems could be invaluable for investigating break-ins. The challenge is that police often don’t have a unified approach to identifying and securing sensor data, as technology evolves far faster than official protocols.
In practice, an officer might arrive at the scene and unplug a camera or device without realising it’s overwriting crucial data in a ring buffer. Or they might hastily turn off power to a connected system that had vital logs in volatile memory. Clear guidelines for preserving data are vital. In some scenarios, you must disconnect hardware at once to avoid losing evidence to overwrites. In others, you dare not shut it off or you risk losing critical sensor details.
Detecting Devices on Site
To help investigators, the project is testing ways to spot which devices may be present. One approach uses AI-driven image recognition: an officer could photograph each room, and the algorithm identifies visible motion sensors, cameras, or thermostats. Problems arise when hardware is discreetly designed or disguised, and the continual flood of new gadgets doesn’t help the algorithm keep up.
Another strategy is radio-frequency scanning. With a standard WLAN “sniffer,” police might see which devices connect to the router. A Zigbee sniffer can detect communications between smart bulbs or motion sensors. Ultimately, the idea is that an officer can run these scans on-site, be alerted to possible devices of interest, and be guided on how to preserve their logs or data.
Privacy Constraints and Evidence Quality
For now, the researchers assume a simplified scenario where the burglary victim consents to handing over sensor data in an effort to find the perpetrator. But in reality, data access is governed by privacy law. Without homeowner consent, any data gleaned from these devices would need proper judicial authorisation. The question arises, too, whether the data might have been tampered with—someone intentionally creating false sensor readings, for instance. That’s why a key aspect of ongoing research is how to cross-verify sensor data from multiple sources to ensure consistency.
In Wolfenbüttel, researchers are assembling a test “smart flat” with extensive home technology, exploring which minimal sensor setups yield the most useful evidence. For example, a single air-quality sensor in the kitchen might still reveal a break-in in the living room, if a broken window or sudden occupancy changes air flow or CO₂ levels. The aim is to learn precisely how reliable each piece of sensor data can be, as well as whether logs can be faked or intentionally misled.
The bigger vision is an enduring forensics lab that tracks these ever-updating consumer tech lines. It would also help locate software vulnerabilities or security flaws. Without ongoing analysis, the next generation of appliances could easily slip past investigators, leaving them perpetually on the back foot.
Thanks to the rise in connected technology, policing increasingly requires knowledge of each new gadget that might capture or transmit data. With the right guidelines, plus reliable technical detection tools, investigators can gather evidence that helps confirm or disprove suspects’ accounts. Burglars and other criminals, meanwhile, may try to exploit or spoof these systems. The cat-and-mouse game continues—but the data footprints that modern devices leave behind might just tip the scales in law enforcement’s favour.