The Internet of Things is changing the world. Its great potential depends on the combination of ultra-low-power network edge smart devices and cloud computing to pattern the massive amounts of data to generate useful information. There are two major factors contributing to the rise of the Internet of Things. One is the emergence of high-performance processor chips, and the other is that the manufacturing cost and power consumption of intelligent network edge devices can be very low. These two factors make the universal deployment of the Internet of Things both technically and economically feasible.

Network and big data are key aspects of the Internet of Things that distinguish between common remote monitoring and control. By changing the way of detecting and responding to one or two variables individually, and then analyzing multiple data channels to detect trends and determine appropriate responses, the Internet of Things can make a difference in protecting the environment, improving performance and changing daily life. potential.
In the automotive industry, it is not uncommon for leading manufacturers to start using sensor information collected from a large number of vehicles to improve customer service and new product development. In other consumer goods markets such as home appliances, leading manufacturers are beginning to use the power of the Internet of Things to solve problematic problems by collecting data on customer devices to improve product performance and performance. In the building services sector, data collected from the global elevator and escalator customer base and integrated into the cloud IoT platform is expected to help improve product maintenance and future product design.

There are many other scenarios where you can combine the detected data, including:
Environmental testing, such as mine gas detection to improve workplace safety.

Proximity sensors on the road and onboard acceleration and altitude sensors (automatic driving and accident avoidance).

Personnel detection sensors in the hotel room (without infringement of privacy), to prevent staff from invading the tenants when providing room service, and to improve operational efficiency.

Patient and environmental data are recorded for transmission to medical sensors of healthcare professionals.

In-vehicle communication that enables automotive data recording to determine insurance rates based on driving habits and motivate safer driving behavior.

Multi-sensor solution needs and development

By integrating sensors and auxiliary electronics, you can save money and simplify installation when you need to monitor multiple variables simultaneously using sensors. A highly integrated sensor evaluation platform helps develop multi-sensor smart products that can be connected to the Internet of Things at any time.

Arduino is one of the environments designed to simplify the development of multi-sensor solutions. For example, the Arduino Lucky Shield is an expansion board that is compatible with all 5 V and 3.3 V standard Arduino boards. It combines sensors for detecting atmospheric pressure, relative altitude, brightness, temperature, motion and presence. These sensors are housed in a compact 68.6 mm x 53.4 mm board.

It's easier to start with Arduino Lucky Shield because Arduino.org offers several tutorials, including a weather station application that shows how to read the temperature, humidity, and pressure sensor outputs and send them to the OLED display.

ST's X-NUCLEO-IKS01A2 board and SensorTile
STMicroelectronics' STM32 ecosystem includes several multi-sensor evaluation boards. The X-NUCLEO-IKS01A2 is an environmental test expansion board for use with the STM32 Nucleo microcontroller substrate. It consists of a MEMS accelerometer, a gyroscope, a magnetometer, an absolute air pressure sensor, and a capacitive relative humidity and temperature sensor.

The STM32Cube ecosystem provides tools and software for initializing and running STM32 microcontrollers. In addition, the X-CUBE-MEMS1 Environmental Sensor Software Extension Library provides the drivers required to build applications based on the X-NUCLEO-IKS01A2. In the overall system architecture diagram shown in Figure 3, X-CUBE-MEMS1 is used to meet the requirements of the driver layer.

In conjunction with the middleware layer shown in Figure 3, other software examples are provided that use sensors for specific functions such as activity and/or gesture recognition. These include:

osxMotionAW: Real-time wrist activity recognition software extension for STM32Cube

osxMotionID: Real-time motion intensity detection software extension for STM32Cube

osxMotionFX: Real-time sensor fusion software extension for STM32Cube

osxMotionGC: Real-time gyroscope calibration software for STM32Cube

osxMotionPE: Real-time pose estimation software extension for STM32Cube

The pseudocode in Figure 4 shows how MotionFX implements real-time motion sensor data fusion.

Pseudo code sequence initialization (will be executed once)

Initialize the sensor (6x fused accelerometer and gyroscope, and 9x fused magnetometer); wait for transients after power up to get good data samples

Initialize MotionFX Fusion: osx_MotionFX_initialize()

Initialize magnetometer calibration:

osx_MotionFX_compass_Init()

osx_MotionFX_getKnobs();

modify settings; _setKnobs()

Reset by disabling fusion: osx_MotionFX_enable_6X(0) / _9X(0)

Start fusion

Initialize gyroscope calibration if possible: osx_MotionFX_setGbias()

Initialize the magnetometer calibration if possible: osx_MotionFX_compass_setCalibrationData()

Enable data fusion: osx_MotionFX_enable_6X(1) / _9X(1)

Sensor data can then be read and commands can be controlled using commands such as osx_MotionFX_propagate() and osx_MotionFX_update()

Small form factor IoT lab

ST recently released a smaller multi-sensor module that can be used as a detection and connectivity hub in embedded systems or as a stand-alone device for acquiring sensor data through smartphone applications. This SensorTile integrates a MEMS accelerometer, gyroscope, magnetometer, absolute pressure sensor and microphone, and an STM32L4 microcontroller and Bluetooth low energy (BLE) on a board that is similar in size to a stamp that can be soldered or plugged into a motherboard. ) Radio.

To be used in stand-alone mode, the ST offers a saddle-shaped board that contains additional temperature and humidity sensors and can be easily modified to add replacement sensors when needed. When used in this mode, SensorTile can be configured via BLE to quickly get sensor data on the smartphone.

For embedded development, the SensorTile can be plugged into the STM32 Nucleo evaluation board via a different extended saddle board.

Samsung ARTIK platform with enterprise-class security

Samsung's ARTIKTM platform offers a range of modules that can be expanded from small units with ARM® Cortex®-M4 microcontrollers and Bluetooth 4.2 support to dual Cortex-A7 processing and support for Bluetooth, Wi-Fi, ZigBee® And Thread's ARTIK 5 series, and the ARTIK 7 series with the Cortex-A35 application processor. The ARTIK 5 and ARTIK 7 series are very powerful and fully available for gateways or controllers. They feature enterprise-class security, including a hardware security element that can be used for cryptographic algorithms and secure OS key storage and secure execution to help build a trusted execution environment. Many premium brands have built the IoT solution using the ARTIK ecosystem, while providing kits for embedded development, such as the ARTIK 020 Bluetooth 4.2 IoT Terminal Kit, ARTIK 520 Bluetooth/Wi-Fi/ZigBee/Thread Kit, and the high-end ARTIK 710 Kit. Fast multi-sensor development is available with ARTIK sensor expansion boards that are compatible with ARTIK 5 and ARTIK 7 kits. The board contains an accelerometer, gyroscope, humidity sensor, magnetometer, pressure and temperature sensor, and is connected to the main evaluation board via an edge connector as a companion module

To sum up

The sensor development boards currently on the market are compact multi-sensor modules that can be used in the final product for IoT edge applications, either directly or with minimal customization. As the surge in user demand and cloud-based analytics applications become more sophisticated and economical, creative services will continue to emerge in the market, and the use of various sensor data will continue to increase.