About
MonadCount is Jakub Dubec’s PhD project at FIIT STU Bratislava: counting the people in a room from the Wi-Fi and Bluetooth signals that are already there.
MonadCount investigates how Wi-Fi measurements change when people move through a room, and whether we can use them to estimate occupancy as the room changes. The work combines measurements in the faculty library with simulation.
This site explains the equipment and publishes the experiments. The map shows where the radios are, the experiment register records the questions and what came back, and the fleet shows what every radio is doing this minute.
Wi-Fi signals reach a receiver along several paths at once: straight across, and bounced off walls, desks and shelves. People bend those paths too. The radios record Channel State Information (CSI), which describes how a signal arrived across the Wi-Fi channel.
The library equipment uses Raspberry Pi 5 computers with Intel AX210 radios. During a measurement, one box transmits frames and the others record what arrives. The research asks which changes come from people, which come from the room, and when a model needs recalibrating.
Bluetooth Low Energy (BLE) monitoring is already active. It gives a second view of the radio activity in the space. We are testing how much that view adds to the Wi-Fi measurements.
The sensing boxes have no cameras or microphones. CSI measures a radio channel, and people affect it whether or not they carry a phone. BLE monitoring observes Bluetooth transmissions from nearby devices.
Volunteer app sessions are a separate activity. They may involve walking a marked route or standing at a marked point, and the phone's measurements are then compared with the radio readings.
See the data protection page for the recording notice and contact details. Questions about a particular measurement are welcome by email.
Jakub Dubec, PhD student at the Faculty of Informatics and Information Technologies, Slovak University of Technology in Bratislava.
We build the software and the measurement equipment, run the experiments, and write up what they tell us. The question looked smaller before we started buying antennas.
Students can ask about the upcoming volunteer beta or talk to us about a bachelor’s or master’s thesis. You do not need prior experience with wireless sensing to start a conversation.
Researchers working on CSI sensing, calibration or crowd modelling are welcome to get in touch too. The code and research vault are on GitHub.
Browse the experiment register for research questions and findings, or inspect simulation runs and their trajectory replays.
The research vault and the code behind this website are open on GitHub, dead ends included.
Last updated 2026-09-28.