In this project we have implemented Omron's D6T-44L MEMS Thermal sensor using STM32L-Discovery board. These are high sensitivity sensors that enable detection of stationary human presence.These are high-precision area temperature detection with low cross-talk field of view characteristics along with superior noise performance and digital output. In this project the sensor was installed at a location and continuously provided the temperature readings. These readings were used and based on a threshold level we could assert whether a person was present or not.
Processor Requirements
Main features of the board useful for the project are as follows.The datasheet can be accessed here[1]
This is a high sensitivity sensor that detects the surface temperature of a radiating body. This is different from traditional pyroelectric sensors which detect only change of signal (i.e. moving bodies, not the stationary one).The D6T series sensors are made up of a cap with silicon lens, MEMS thermopile sensor chips, and dedicated analog circuit and a logic circuit for converting to a digital temperature value on a single board through one connector.Datasheet can be found here[2].
This sensor works on the Seebeck Effect[3].The silicon lens collects radiated heat (far-infrared ray) emitted from an object onto the thermopile sensor in the module.The radiated heat (far-infrared ray) produces an electromotive force on the thermopile [4]sensor.The analog circuit calculates the temperature of an object by using the electromotive force value and a measured temperature value inside the module.The measured value is outputted through an I2C bus which can be used for further processing using STM32L-Discovery Board.
D6T-44L-06 has sensor chip arrays of 16 channels (4x4).Each channel corresponds to a pixel (as shown in the figure below) and measures temperature independently. In the Field of view of the sensor that includes all the pixels, whenever an object appears the temperature of corresponding pixel changes.By mounting the signal processing circuit closely to the sensor chip,a low noise temperature measurement is realized.The application note can be seen here.[5][6]
Please note that the sensitivity area is wider than the FOV specified area. When an object to be measured is smaller than the sensitivity area, the background temperature effects the measurements. Though Omron’s D6T sensor corrects a temperature measurement value by using a reference heat source (blackbody furnace), the measurement’s value is influenced by the emissivity of the specific material of the object to be measured, and the surface shape of the occupant relative to the sensitivity area.
Note that the occupied area in FOV becomes smaller with increasing distance and the background temperature prevails. So for detecting human beings, the application will be limited to close range when the detection programming scheme only judges by temperature value. To extend the detection distance, improvements to the judgment accuracy can be made via software programming, considering time change, heat source location and human being movement.
This project can be combined with the "Intelligent Load Balancing under constrained energy availability"[7] so as to add on to home automation. This sensor can be used with the spark core wifi enabled board to include the above mentioned two projects in one board. As we are controlling this over the internet, cyber security should also be provided.