A simple Internet connection and a small electronic card priced at 10 euros can be used to create a motion detector that lights up a corridor when someone walks by. This technique uses a free software named ESPectre, which turns an ESP32 card into a motion sensor. The ESP32 is a microchip produced by the Chinese company Espressif, featuring built-in Wi-Fi and widely used by enthusiasts of home automation. ESPectre was developed by Francesco Pace, a mathematician and informatician with over 20 years of experience. He introduced his project in January 2026 on IoT For All, and it gained more than 4,000 stars on GitHub at that time, now reaching over 9,000. Wi-Fi signals travel through multiple paths simultaneously between the router and the ESP32. Some signals go directly, while others bounce off walls, furniture, and people. The human body, which is about 60% water, absorbs and reflects waves at 2.4 GHz, the frequency used by most Wi-Fi devices. When someone moves, these signal paths change. To detect these changes, ESPectre reads the Channel State Information (CSI), which describes the state of the radio channel. Wi-Fi splits its signal into several sub-frequencies (64 on the ESP32), and for each one, the chip measures the signal strength and its shift. In an empty room, these values change very little. They fluctuate when someone walks through. The software continuously calculates the magnitude of these fluctuations and compares them to a threshold set at startup. Francesco Pace uses no artificial intelligence, only statistical methods and filters. The principle is similar to moving a hand in front of a flashlight and watching the shadow shift. A passive infrared sensor (PIR), used in motion lamps, must "see" the person. Wi-Fi, however, can pass through walls made of plaster. A single card can monitor a neighboring room of about 50 square meters, according to the developer. However, ESPectre does not detect someone who remains still. It also cannot distinguish between a human and a cat. It is important to choose the right card, as not all chips provide access to the CSI. Francesco Pace recommends the ESP32-S3 or the ESP32-C6. The ESP32-C3, which is cheaper, also works, as does the original ESP32, though it is less optimized. The ESP8266, an older version of the ESP32 used in many smart plugs, does not provide access to the CSI. A development card costs about ten euros. You also need a USB cable that transfers data (charging-only cables are a common mistake). A Home Assistant server, which can run on a Raspberry Pi costing 35-50 euros or on a PC that is always on, is also needed. You can even get help from an AI to install Home Assistant. On the other hand, a Philips Hue motion detector costs 39.99 euros. The installation involves plugging in the ESP32 via USB and opening the ESPectre installer in Chrome on a computer. Firefox and Safari do not support flashing the card, which means copying the software onto it. Select the card's port, and the assistant detects the chip. Then choose the ESPHome version, the one that connects to Home Assistant. Enter your 2.4 GHz Wi-Fi network. Place the card between 3 and 8 meters from the router, at table height. Keep it away from the refrigerator and metal furniture. At each startup, remain still for 10 seconds, allowing for calibration. In Home Assistant, the card appears automatically, with a sensor labeled "motion detected." Then create an automation to turn on a light or an outlet. You can also receive an alert when you are away, like with a smart alarm. At XDA, the writer placed his ESP32-S3 about 3 meters from his router and activated the "high precision" profile. He found the detection surprisingly accurate. He does not, however, use it for the living room, where he would quickly be detected. His setup is part of other low-cost home automation projects, such as a 15 euro AliExpress screen or an old e-reader converted into a dashboard. While AI is not required for these setups, it can assist with the most delicate steps and technical choices. The card does not capture images or sounds, making it less intrusive than a surveillance camera. However, the developer reminds users that these motion data can reveal the habits of a household. Therefore, it is necessary to obtain the consent of the people being monitored. Inform those who live under your roof before plugging in the card.