PMV sensor stations



Duration from 05/2016
Responsible Prof. Dr.-Ing. Sabine Hoffmann
Team Mathias Kimmling, Konrad Lauenroth, Sabine Hoffmann
Project type hardware development

The Internet of Things and the emerging of inexpensive hardware such as sensors and microcontrollers offers new opportunities in building automation. To determine thermal comfort or discomfort at workplaces, a PMV sensor station was developed which could be deployed comprehensively in office buildings because of its low investment cost.

The PMV-index (Predicted Mean Vote) describes how the population mean would be likely to rate their thermal sensation under given ambient room conditions. The index can thus serve as a quality criterion for the thermal situation in a building. PMV can take on values from -3 (cold) to +3 (hot), with 0 being a neutral state, i.e. neither too cold nor too hot. The sensor station measures the values required for PMV-calculation by using inexpensive sensors for air temperature, relative humidity, air speed, and globe temperature.

A microcontroller (ESP32/NodeMCU) is the platform for connecting the sensors, controlling the measurement intervals, recording the measurement data, and transferring them via Wi-Fi to a MariaDB database server. In addition, automatic measurement value checking, and error handling is implemented in the software of the sensor station. Together with user input on physical activity and clothing factors, the PMV index can be calculated on the database server.

By comparing the PMV values with additionally queried, subjective voting on thermal comfort, systematic measurement errors can be identified, and personal preferences recognized. For this purpose, a graphical real-time display of the measured values and calculated PMV via a Grafana platform is also available.

A network of several PMV sensor stations is used to record the indoor environmental situation in the Living Lab smart office space on an individual workstation basis and thus provides an important data basis for ongoing building simulations and the control of devices in the Personalized Environment