December 29 is the last working day of 2023, in today's "first share of humanoid robots" - UBTECH, officially listed on the main board of the Hong Kong Stock Exchange, with 9980 as the stock code, UBTECH sold 11.282 million H shares globally, with a final issue price of 90 Hong Kong dollars, raising 1.015 billion Hong Kong dollars, net raising of about 906 million Hong Kong dollars, the first day of listing The stock was stable, and the market value was about 37.6 billion as of press time.
Sensor Expert Network
https://www.sensorexpert.com.cn
Focusing on the field of sensor technology, it is committed to providing professional and vertical services for global cutting-edge market trends, technology trends and product selection, and is a leading sensor product query and media information service platform in China. Based on sensor products and technologies, it provides accurate matching and docking for the majority of electronic manufacturing practitioners and sensor manufacturers.
It was a historic moment when the humanoid robot company gained market recognition and landed on the financial markets. The outbreak of humanoid robots in 2023 has also driven the development of downstream domestic components, and its Chinese-made sensor industry is highly favored, and mid-to-high-end sensing companies represented by six-dimensional force sensors, flexible sensors, image sensors, and MEMS sensors have ushered in a wave of investment and financing boom.
This article comes from a paper published by You Zheng, an academician of the Chinese Academy of Engineering, a few years ago, about the research and development of intelligent sensor technology. The article predicts several major application areas for smart sensors in the future, including robotics.
Academician You Zheng believes that China should seize the historical opportunity brought by the intelligent era to the development of the sensor industry, comprehensively improve the basic research and industrialization level of intelligent sensors, and provide strong technical support for the arrival of the intelligent era.
Nowadays, technologies such as humanoid robots and intelligent driving are booming, especially relying on sensor technology!
Original title of the paper: Research Progress and Application Prospect of Intelligent Sensor Technology
Author: Academician of the Chinese Academy of Engineering | You Zheng
Smart sensor refers to a sensor with information detection, information processing, information memory, logical thinking and judgment functions.
Compared with traditional sensors that only provide analog voltage signals to represent the physical quantities to be measured, smart sensors make full use of integrated technology and microprocessor technology to integrate perception, information processing, and communication, and can provide information with a certain level of knowledge transmitted in digital mode.
Since NASA put forward the concept of smart sensors in the 80s of the 20th century, after decades of development, smart sensors have become a major development direction of sensor technology, representing a country's industrial and technological research capabilities.
Driven by the current intelligent era, the importance of sensors has become more prominent, not only in the "Made in China 2025", "Germany 2020 High-tech Strategy" and the European Union, the United States, South Korea, Singapore and other smart city strategies to play an important supporting role, but also in the Internet of Things, virtual reality (VR), robots, smart home, autonomous vehicles and other industries play a key role in the development of the industry.
The rise and development of high-performance, high-reliability multi-functional and complex automatic, measurement and control systems, and the Internet of Things based on radio frequency identification technology have increasingly highlighted the importance of intelligent sensors with perception and cognitive capabilities and the urgency of their vigorous and rapid development.
With the development of CMOS-compatible MEMS technology, the development of miniature smart sensors has received strong technical support, and the smart sensor industry is facing a very important historical development opportunity.
In this paper, the development status of different types of smart sensor technologies and applications is reviewed, and the future development trends are prospected.
1 A wide variety of smart sensors
In order to meet the needs of various intelligent applications, the sensor categories are very diverse, such as: environmental sensors, inertial sensors, analog sensors, magnetic sensors, biological sensors, infrared sensors, vibration sensors, pressure sensors, ultrasonic sensors, etc.
Among them, the following sensors are more commonly used.
Environmental sensors, mainly gas sensors, barometric pressure sensors, temperature sensors, humidity sensors, etc. Gas sensors can be used in air purifiers, drunk driving monitors, detectors of toxic gases such as formaldehyde in home improvement, and detection devices for industrial exhaust gases. With people's attention to environmental issues, the importance of environmental sensors is becoming more and more prominent, and there is a lot of room for development in the future.
Inertial sensors are mainly used in wearable products, such as smart bracelets, smart watches, VR helmets, etc. Through the inertial sensor, it detects the tracking and recognition of movement, and informs the wearer of the amount of exercise that day, the calories burned and the effect of exercise.
Magnetic sensors, mainly used in household appliances, such as coffee machines, water heaters, air conditioners, etc., are used to detect how much the angle has been turned or how much travel has been made, and are usually displayed on the dashboard. In addition, magnetic sensors are also used in door magnetism and window magnetism, and the intelligence and accuracy of the robot also need magnetic sensors to support.
Analog sensors, mainly used in smart medical equipment, can be used as the input of heartbeat, electrocardiogram and other signals, and the output of health data is visualized, so that users can understand their first-hand health and exercise data.
Infrared sensors are often used in smart homes such as infrared cameras and sweeping robots.
2 Technological research progress of smart sensors
A truly smart sensor should have the following functions:
1) Self-calibration, self-calibration and automatic compensation function;
2) Automatic data collection, logical judgment and data processing functions;
3) Self-adjusting and adaptive function;
4) a certain degree of storage, identification and information processing functions;
5) Two-way communication, standard digital output or symbol output function;
6) The function of algorithm judgment and decision-making processing.
The following takes commonly used temperature, pressure, inertial, biochemical and RFID sensors as examples to introduce the research progress of intelligent sensing technology.
2.1 Smart Temperature Sensor
The development of temperature sensors has roughly gone through the following three stages: traditional discrete temperature sensors, analog integrated temperature sensors, and intelligent temperature sensors.
After entering the 21st century, intelligent temperature sensors are developing rapidly in the direction of high precision, multi-function, bus standardization, high reliability and security, the development of virtual sensors and network sensors, and the development of monolithic temperature measurement systems.
Today's smart temperature sensors consist of temperature sensors, A/D converters, signal processors, memory, and interface circuitry, and some include multiple selectors, central controllers, random access memory, and read-only memory.
The characteristics of the intelligent temperature sensor are that it can output temperature data and related temperature control quantities, adapt to various microcontrollers, and realize the test function through software on the basis of hardware, and its intelligence depends on the level of software development.
1) Improve measurement accuracy and resolution
The earliest intelligent temperature sensors began in the mid-90s of the 20th century, using 8-bit A/D converters, which have low temperature measurement accuracy and a resolution of only 1°C.
At present, a variety of high-precision, high-resolution intelligent temperature sensors have been launched abroad, using 9~12-bit A/D converters, and the resolution can reach 0.5~0.625°C. The DS1624 high-resolution intelligent temperature sensor, newly developed by Dallas Semiconductor in the United States, can output 13-bit binary data with a resolution of up to 0.03°C and a temperature measurement accuracy of ±0.2°C.
In order to improve the conversion rate of multi-channel intelligent temperature sensors, some chips use high-speed successive approximation A/D converters. For example, the AD7817 5-channel intelligent temperature sensor has a conversion time of only 27 ms for the local sensor and 9 ms for each remote sensor.
In terms of high-precision temperature measurement, some scholars have designed a high-performance digital temperature sensor, which is composed of a quartz tuning fork resonator, a digital interface circuit and a sensor reset control algorithm based on a field-programmable gate array, and the sensitivity of the sensor can reach the number of 10 -6 °C, that is, the temperature measurement resolution is 0.001 °C, the response time is 1 s, and the measurement accuracy is 0.01 °C.
2) Enhanced testing capabilities
The testing capabilities of the new smart temperature sensors continue to increase. The intelligent temperature sensors have a variety of working modes to choose from, mainly including single conversion mode, continuous conversion mode, standby mode, and some also add low temperature limit extension mode.
For some smart temperature sensors, the host computer (external microprocessor or microcontroller) can also set its A/D slew rate, resolution, and maximum conversion time via the corresponding registers.
In addition, intelligent temperature sensors are developing from single channel to multi-channel, which creates good conditions for the development of multi-channel temperature measurement and control systems.
3) Standardization and normalization of bus technology
At present, the bus technology of intelligent temperature sensors has also been standardized and standardized, and the buses used mainly include single-wire (-Wire) bus, I 2 C bus, SMBus bus and SPI bus.
4) Reliability and safety design
In order to avoid malfunction when the temperature control system is disturbed by noise, a programmable fault queue counter is set inside some intelligent temperature sensors, which is dedicated to setting the number of times the measured temperature value exceeds the upper and lower limits. The interrupt port can only be triggered when the measured temperature continuously exceeds the upper limit or falls below the lower limit for a set number of times, avoiding the impact of accidental noise interference on the temperature control system.
In order to prevent damage to the chip due to human electrostatic discharge, some intelligent temperature sensors also add electrostatic protection circuits, which can generally withstand an electrostatic discharge voltage of 1~4 kV.
For example, the TCN75 smart temperature sensor can withstand an electrostatic discharge voltage of 1 kV on the serial interface, interrupt/compare signal output, and address input. The LM83 intelligent temperature sensor withstands an electrostatic discharge voltage of 4 kV.
2.2 Smart Pressure Sensors
Intelligent pressure sensors are the combination of microprocessors and pressure sensors, so their implementation methods can be divided into: non-integrated intelligent pressure sensors, integrated intelligent pressure sensors and hybrid intelligent pressure sensors.
The non-integrated intelligent pressure sensor is an intelligent pressure sensor system that combines the traditional pressure sensor, signal conditioning circuit, and microprocessor with digital bus interface.
This non-integrated pressure sensor is actually an addition of a microprocessor connection to a conventional pressure sensor system. Therefore, it is the fastest way and way to achieve an intelligent pressure sensor system.
The integrated intelligent pressure sensor is a monolithic integration of pressure-sensitive components with signal processing, calibration, compensation, microcontroller, etc., mainly using microelectromechanical system (MEMS) technology and large-scale integrated circuit process technology, using silicon as the matrix material to make sensitive components, signal conditioning circuits, microprocessing units, and integrate them on a chip.
With the rapid development of microelectronics technology and the application of micro and nano technology, the intelligent pressure sensor made by this has the characteristics of miniaturization, structural integration, high precision, multi-function, array, full digitalization, easy to use and simple operation.
The hybrid intelligent pressure sensor is to integrate the various integrated links of the system, such as the sensitive unit, signal conditioning circuit, microprocessor unit, and digital bus interface, in different combinations on 2~3 chips and package them in a shell.
Hybrid integration to achieve intelligence is an intelligent approach that is very suitable for the current technology development. In an intelligent pressure sensor system, a microprocessor can implement software control of the sensor according to a given program, turning the sensor from a single function to a multi-function. Smart pressure sensors generally have the following basic functions.
1) Data processing function. The intelligent pressure sensor not only measures each measured parameter, but also can automatically zero, auto-balance, and auto-compensate according to the known measured parameters.
2) Automatic diagnosis function. This is the main function of the intelligent pressure sensor, the intelligent pressure sensor through its fault diagnosis software and self-detection software, automatically on the sensor and system working status of regular and irregular detection, testing, timely detection of faults, assist in diagnosing the cause of the fault, location, and give operation tips.
3) Software configuration function. Due to the use of microprocessors, intelligent pressure sensors not only have the necessary hardware components, such as detection, amplification, A/D, D/A, communication interfaces, etc., but also have software resources for controlling and processing data. In the intelligent pressure sensor, there are multiple modular hardware and software, and the user can send commands through the microprocessor to complete different functions, which increases the flexibility and reliability of the sensor.
2.3 Intelligent Inertial Sensors
Inertial sensors are the most widely used type of MEMS sensors, including accelerometers, gyroscopes, and azimuth sensors. MEMS technology has the unique advantage of enabling the miniaturization and cost reduction of inertial sensors.
Today's inertial measurement units (IMUs) can integrate a 3-axis accelerometer, a 3-axis gyroscope, and a 3-axis magnetometer in a size of 10 mm×10 mm, × 4 mm at a cost of less than $1. This inertial measurement module can be applied to smart phones and wearable devices to realize functions such as gait monitoring, step counting, fall detection, sleep monitoring, indoor navigation and other sports and health functions, as well as gesture recognition, direction perception and other entertainment functions.
1) Smaller, more flexible, more energy-saving, high-performance, and highly integrated.
Intelligent inertial sensors used in wearable devices need to have smaller size, lower power consumption, and wireless data transmission as a node of the body area network, and finally achieve flexibility.
The world's smallest three-axis accelerometer is the BMA355 released by Bosch in 2014, which is packaged in a wafer-level package with dimensions of only 1.2 mm× 1.5 mm× 0.8 mm, and extremely low power consumption, operating at only 130 μA, which can be reduced to 1/10 in low-power mode.
In addition, the BMA355 also has a powerful intelligent terminal engine, and the interrupt modes include data ready synchronization, motion wake-up, tap sensing, direction recognition, horizontal and vertical toggle switches, low-g/high-g-value impact detection, free-fall detection, power saving management, etc., which can be used in wearable devices such as health trackers, pedometers (smart watches and bracelets), jewelry, etc.
In addition to the application of wearable devices, inertial sensors also have broad applications and development prospects in the military field, which are different from the requirements of wearable devices, and military applications put forward higher requirements for sensor accuracy, reliability and stability under extreme conditions.
Inertial sensors use the inertia of the mass to measure the mass to be measured, while the MEMS sensor has a small mass, taking the gyroscope as an example, its accuracy is generally not as good as that of the traditional gyroscope, and it is difficult to be directly applied in high-end fields such as aviation and aerospace.
According to the current level of technology, the accuracy of a single MEMS gyroscope is close to the limit of the current stage, and new methods are needed to improve the accuracy of MEMS gyroscopes.
2) Multi-sensor integration and data fusion.
Considering the small size and low cost of MEMS sensors, multi-sensor integration and data fusion technology can be used to improve accuracy, that is, the performance of a single sensor can be achieved through the information fusion of multiple sensors.
NASA put forward the concept of virtual gyroscope in 2003, that is, the use of multiple MEMS gyroscopes to form an array, redundant detection of the same signal and output multiple detection values, the use of data fusion technology to analyze and synthesize these detection values, the gyro array is fused into a virtual gyroscope, and the optimal estimate of the input angular rate is obtained, which greatly improves the accuracy of the gyroscope.
Subsequently, the micro-nano laboratory of Northwestern Polytechnical University filtered three microgyroscopes with bias stability of 35.00 (°)/h, and the drift performance of the virtual gyroscopes was improved by more than 200 times, which demonstrated the generalization of virtual gyroscopes
The feasibility of the concept also provides a new method and new idea for using array sensors to improve accuracy.
3) New sensitive mechanisms.
Another way to improve the performance of existing MEMS sensors is to discover new sensitive mechanisms. In 2015, the micro-nano laboratory of Northwestern Polytechnical University demonstrated the world's first resonant accelerometer based on modal localization. The traditional resonant accelerometer is changed by detecting the resonant frequency change sensitive acceleration, but by detecting the change of the amplitude ratio of two weakly coupled resonators, the sensitivity is increased by 300 times, which opens up a new way for the development of high-precision inertial sensors.
At the same time, based on the phenomenon of forced thermal convection, the research group designed a multi-axis inertial sensor "ejector roton gyroscope", which can be sensitive to angular velocity in up to three directions and linear acceleration in three directions at the same time. The use of fluid particles instead of solid masses also opens up a relatively new field of MEMS research.
Fluid inertial sensing omits movable parts, and has the characteristics of simple structure and high stability of the device. With the new discovery of sensitive mechanisms, the development of microelectromechanical technology and the application of new materials, MEMS inertial sensors will further diversify and refine their types, and play a more important role in consumer electronics such as wearable devices, inertial navigation and automatic control.
2.4 Radio Frequency Identification Technology
Radio Frequency Identification (RFID) is a communication technology that uses radio signals to automatically identify specific targets and read and write relevant data, without the need to establish mechanical or optical contact between the identification system and the specific target. According to whether the label is active or not, it can be divided into passive label, semi-passive label and active label.
Passive tags, also known as passive tags, derive energy from the RFID reader's interrogating radio waves. Active tags, also known as active tags and semi-passive tags, have an internal power supply that can be identified at a distance of several hundred meters from the RFID reader.
The difference between the two is that an active tag can emit a signal without the energy provided by the reader, while a semi-passive tag still relies on the energy emission signal provided by the reader. In contrast to barcodes, RFID tags do not need to be in the reader's line of sight when they are recognized, so RFID technology can be embedded in the identified object.
RFID is a method of intelligent identification and data acquisition (AIDC) and an important part of the Internet of Things (IoT), mainly used in defense and security, identification, environment, transportation, medical health, agriculture and animal husbandry and other fields.
The core technology of RFID includes RFID antenna technology, data integrity and security, RFID middleware technology and RFID standard system. In recent years, RFID research hotspots have mainly focused on data integrity and security, such as ensuring that users' privacy is not leaked while obtaining information, privacy protection when the ownership of items containing RFID tags is changed, and the use of RFID technology to achieve applications in other fields, such as indoor positioning based on RFID technology.
2.5 Intelligent biochemical sensors
Biochemical sensor refers to a device that can sense biochemical quantity and convert it into useful signal output according to a certain law, which is generally composed of two parts:
With the development of materials science, the biochemical sensitive membrane formed by two-dimensional new materials has shown more superior performance, and has gradually become a hot spot in the research field of biochemical molecular recognition components.
The second is a signal converter, which is mainly composed of electrochemical or optical detection components, such as current potential measurement electrodes, ion sensitive FETs, etc.
With the continuous development of new materials, new principles and new integration technologies, especially the emergence of MEMS technology and biochip technology, the research of biochemical sensors has gradually developed into the research of biochemical systems characterized by miniaturization, integration and intelligence.
In the past, sensor research only focused on improving its own performance, such as sensitivity, dynamic range, response time, reliability, etc., but with the continuous integration of MEMS technology and standard CMOS technology, the integration of sensors and readout circuits has become possible, and with the continuous progress of hybrid integration technology, more functional circuits, including communication modules, energy harvesting, and power management modules are integrated into intelligent biochemical sensors, laying a technical foundation for the miniaturization, multi-functionality and intelligence of sensors.
In order to truly realize the miniaturization and intelligence of sensors, biosensors need to be integrated with active circuits to form a multi-functional system-on-chip.
With the continuous development of new materials, new structures and new principles, DNA sensors based on cantilever beams, protein/DNA sensors based on polycrystalline silicon nanowires, blood glucose sensors based on hydrogels, pH sensors based on ion sensitive FETs and temperature sensors based on bandgap reference can be integrated on the same chip with their corresponding readout circuits, wireless communication and other modules, with self-calibration function, and can achieve self-adjustment and self-adaptation functions within a certain range.
In practical applications, multiple biochemical signals often need to be detected at the same time, which requires a multi-sensor system-on-chip that uses different detection principles to achieve simultaneous detection of multiple signals.
The implementation of multi-sensor system-on-chip presents many challenges for IC back-end process design, because chips with multiple sensors need to go through multiple back-end processes, all processes must be compatible with standard CMOS processes, and back-end processes must also be compatible with each other.
In order to achieve real-time monitoring of multiple physiological parameters, four commonly used sensors in biochemical examinations (including polycrystalline silicon nanowire-based protein sensors, hydrogel-based blood glucose sensors, ion-sensitive field-effect transistors-based pH sensors, and temperature sensors based on bandgap reference) are integrated on the same chip.
In order to realize the miniaturization of intelligent multi-sensor, the analog circuit adopts a reconfigurable multi-sensor interface, a programmable gain amplifier, and a 10-bit SAR ADC structure, which significantly reduces the chip area.
In addition, in order to achieve energy self-sufficiency of smart sensors, two energy harvesting methods are adopted at the same time (including gallium arsenide solar cells to collect light energy and electromagnetic coupling methods to collect radio frequency energy), which solves the problem of replacing batteries in long-term use or implantation application scenarios of medical devices.
The use of reconfigurable circuits to reduce power consumption and the use of energy harvesting methods to extend battery life have also become hot spots in the research of intelligent biochemical sensors.
Figure 2 shows a reconfigurable multi-sensor system-on-chip micrograph, the chip is fabricated using TSMC 0.35 μm CMOS process and the necessary back-end process, the chip area is 3 mm×3.75 mm, and the measured performance parameters are shown in Table 1. The above-mentioned intelligent biochemical sensors and their electronic circuits are still made of hard electronic materials mainly made of silicon materials.
With the development of wearable sensors, lightweight and flexible materials have been gradually used as substrates for the parts of the sensor that are in direct contact with the human skin to eliminate the foreign body sensation of device wearing.
Wearable sensors provide diagnostic and monitoring functions to monitor physiological and biochemical signals as well as motion sensing.
Physiological and biochemical signal monitoring helps to diagnose neurological diseases (such as epilepsy), cardiovascular diseases (such as hypertension), lung diseases (such as asthma), etc., and continuously monitor the treatment process. Continuous monitoring of vital signs, such as heart rate and breathing rate, can provide important data support for early diagnosis and clinical intervention of chronic diseases.
The continuous promotion of this series of wearable sensors also provides a terminal hardware foundation for the construction of future telemedicine diagnosis systems.
At present, wearable smart sensors can realize the extraction of the above multiple physiological and biochemical parameters, and the sensor and its peripheral circuits can be integrated into lightweight flexible substrates, and the 2016 International Conference on Solid-State Circuits (ISSCC) showcased the latest research results in this field.
A new sensor system for oxygen saturation and biopotential signal detection is shown in Figure 3, in which an organic photodiode (OLED), an organic light detector (OPD), a biopotential signal electrode, and a system-on-chip containing body channel communica⁃tion (BCC) circuitry are mixed and integrated on a flexible PET substrate with an overall area of 2.5 cm × 5.5 cm, including a battery mass of 2 g, and a system power consumption of 141 μW.
The optical detection part of the intelligent sensor has a self-calibration loop, the system has automatic data acquisition and data processing capabilities, the blood oxygen saturation data and the electrocardiogram signal data collected by the sensor node are transmitted to the central sensor through the limb communication (BCC) transceiver, in addition, the clock signal is sent to each sensor node by the central sensor, so as to remove the external off-chip crystal oscillator of each node, and the system architecture improves the integration degree of the system while realizing the two-way communication between the sensors.
At present, in the research of wearable sensors, the basic functions of intelligent sensors such as self-calibration, automatic acquisition, and two-way communication can be realized, and how to improve the synchronization accuracy in the process of multi-signal acquisition in complex detection environments has become a research hotspot.
In 2016, the European Research Center for Microelectronics (IMEC) and Samsung Electronics jointly demonstrated a multi-parameter physiological signal recording platform with built-in concurrent electrocardiogram (ECG), bioimpedance (BIO-Z), skin flow response (GSR), and photoplethysmography (PPG) pulse wave sensors to achieve multi-parameter simultaneous acquisition, which can provide more accurate, reliable, and extensive health assessment for wearable electronics.
Because multiple sensors use the same chip for data acquisition, high-precision synchronization can be achieved between data streams, thus providing a basis for correlation analysis and data fusion between multiple data.
For example, ECG and PPG data can be combined to analyze pulse arrival time and further estimate blood pressure values, and ECG, PPG and BIO-Z data can be combined to estimate hemodynamic parameters more accurately.
This multi-parameter synchronous acquisition system can provide more accurate time series data for future data analysis and provide a basis for more biochemical parameter estimation and calculation, which is the research hotspot and inevitable trend of future development of smart sensors.
3 Market applications of smart sensors
In terms of market applications, sensors can not only promote the upgrading of traditional industries, such as the upgrading of traditional industries and the intelligent upgrading of traditional home appliances, but also promote innovative applications, such as robots, VR/AR (virtual reality/augmented reality), drones, smart homes, smart medical care and elderly care.
3.1 Boost to the upgrading of traditional industries
1) Promote the transformation and upgrading of traditional industries.
In the industrial field, traditional enterprises are facing problems such as increasing labor costs and declining market demand, and traditional enterprises have begun to shift from labor-intensive to automated and intelligent. Sensors play a vital role in the whole transformation, helping "Made in China" to "Made in China".
In order to improve the efficiency of the factory, it is necessary to add sensing devices on the production line to track the whole process of products and processes, and at the same time use equipment with sensing devices such as robotic arms and automatic guided vehicle systems to speed up production and accuracy, and improve manufacturing efficiency in an all-round way.
2) Help the intelligent upgrade of the home appliance industry.
In recent years, the performance of household appliance enterprises has declined seriously. According to the relevant data of the Ministry of Industry and Information Technology, in 2015, the total export value of household electronic and electrical products fell by 0.6% year-on-year, and the total import value of household appliances fell by 5% year-on-year.
At present, how to find new growth points and reverse the decline in performance is a major test for the home appliance industry. To this end, traditional home appliance companies have begun to upgrade their home appliances intelligently, and have successively launched smart refrigerators, smart air conditioners, smart washing machines, smart ovens, sweeping robots and other products to meet the personalized needs of users for household appliances.
In terms of the intelligent performance of smart home appliances, for example, smart washing machines can realize the intelligence of washing machines through water level sensors, smart ovens will achieve simple and intelligent baking experience through temperature sensors, etc., and sweeping robots can be supported by adjustable displacement sensors to realize the intelligent and precise operation of robots.
There are many types of home appliances, and in the future, there will be a variety of needs for sensors, such as motion sensors, auditory sensors, vision sensors, microphone arrays, temperature/humidity sensors, etc., which will be used in major appliances and small household appliances.
Therefore, customizable, parameter-adjustable sensors will more effectively support various application scenarios of home appliances.
3) It is expected to bring a turnaround to the mobile phone industry.
As we all know, the global mobile phone industry has entered a state of saturation. GFK, China's smartphone market, predicts that the mobile phone market will grow by only about 3.1% in 2016. Whether the mobile phone industry can usher in a turnaround depends largely on the development of sensors.
In terms of the functions of current smartphones, they are far from meeting people's imagination of mobile phones. With the help of sensors, mobile phones can become more humane and intelligent. For example, smell sensors, taste sensors, and magnetic sensors for truly sensitive motion tracking can all make mobile phones more powerful.
It's safe to say that when the various categories of sensors reach maturity, there will be new opportunities for the mobile phone industry.
3.2 Support for innovative applications
Among the innovative applications of sensors, the most typical are new applications such as robotics, virtual reality/augmented reality (VR/AR), drones, etc.
1) Virtual Reality and Augmented Reality.
Virtual Reality (VR) and Augmented Reality (AR) are among the hottest and most popular applications. The reason why these two technologies are so eye-catching is that VR virtual reality can give people an immersive experience, and AR augmented reality can make people's experience of reality more vivid, intense and intuitive. And these feelings are inseparable from the support of sensors.
In the application of sensors, VR/AR hardware will use nine-axis gyroscopes, infrared positioning sensors, eye-tracking sensors, and gesture recognition sensors, etc., which can obtain information such as the user's movement, posture, and acceleration.
In the future, biosensors will also be used. For example, when children go on a trip, the elderly at home can also get the same experience as their children when they are paired with somatosensory devices with biosensors. VR and AR have been applied to games, sports, education, tourism, cinema, medical and other fields. With the continuous expansion of VR and AR applications, the application demand for sensors will be huge.
2) Robots.
Robot is a programmable and multi-functional manipulator, or a special system that can be changed and programmable by computer in order to perform different tasks, generally composed of actuators, driving devices, detection devices and control systems and complex machinery.
Google's AlphaGo's victory over the Go Grandmaster has sparked global attention to robots. In robots, many sensors are required, including the surrounding environment, posture testing, and human-computer interaction.
Robots require a large number of different types of sensors, and high requirements are placed on the performance of the sensors. Robots are widely used, such as pension industry, industry, service industry, education, etc. If the sensor is done well, the robot industry can fly, which will help the development of many fields such as pension and industry
3) Drones.
Unmanned aerial vehicles (UAVs) have developed rapidly in a short period of time.
According to statistics, the shipment of drones will reach 5 million units this year and exceed 10 million units next year. Sensors play a vital role in the highly integrated and intelligent use of drones.
Gyroscope, infrared, ultrasonic, laser, camera, air pressure, geomagnetic and other sensors will be applied to the UAV, so as to realize the smooth control and auxiliary navigation of UAV technology, as well as humanized obstacle avoidance, recognition, tracking and other intelligent control.
4 Future development of smart sensors
4.1 Sensors are moving towards integration
In order to open up a broader development space, MEMS sensors have begun to move towards integration.
At present, some companies have begun to develop integrated sensors, such as integrating microphones with barometric pressure sensors, integrating barometric pressure sensors with temperature and humidity sensors, integrating microphones with temperature and humidity sensors, etc.
There are several advantages to sensor integration: one is to achieve more powerful product functions to meet diverse needs, and the other is cost advantage, one integrated sensor is more cost-effective than two separate sensors. The third is to reduce the size, which can meet the development needs of more wearable smart products.
4.2 Wireless Energy Harvesting
There are many constraints to traditional sensors, the most prominent of which is the way power is delivered. Conventional sensors are mainly powered by batteries or power lines, which are subject to regular maintenance and replacement costs in addition to installation costs.
In addition, the size of wearable products also puts forward higher requirements for the size of the sensor. In this regard, wireless energy harvesting has become the next development direction of sensors.
Wireless energy harvesting technology refers to the technology that converts energy in the environment such as light, kinetic energy, and heat energy into electrical energy to supply power to the system, so as to realize the self-power supply of sensors, so that the sensors can be placed anywhere, and also reduce the cost of replacement and maintenance.
At present, some foreign companies have launched corresponding solutions and said that the sensor can continue to work for more than 10 years.
In the future, with the continuous advancement of applications, sensors will also be combined with artificial intelligence technology, and the sensor will not be a cold device, but will become a more intelligent and warm product.
4.3 Algorithms and Schemes
With the increase of subdivided application requirements, the importance of software algorithms and solutions on top of sensors is becoming more and more prominent. In terms of algorithms, such as the application of biosensors in the medical and health industry. In addition to heart rate, heart load rate, stress, sleep index, etc., ECG algorithms also include medical applications that have passed FDA approval. In addition, sensor-based solutions are being introduced.
Some sensor companies are starting to offer solutions to detect physical health conditions and are partnering with insurance companies.
Specifically, the sensors in the health device can monitor the user's physical condition, and the insurance company will give these health devices to the user, so as to obtain the user's health information, and set the amount of the user's insurance according to the health data, so as to reduce the insurance company's losses and maximize benefits.
4.4 China's sensor industry should seize the historical opportunity
The sensor industry has a high entry threshold and high barriers, and the investment is large and the risk is large. In the field of sensors, the world's countries with originality and large product volume are mainly concentrated in the United States, Germany, Italy and France.
In contrast, there are some deficiencies in China's sensor industry: there is a serious lack of core sensor technology accumulation such as materials, design, and technology. The scale of MEMS enterprises is relatively small, and the annual sales of MEMS enterprises with complete core independent design and IP do not exceed 100 million US dollars, the maturity of the industrial chain at the MEMS manufacturing end is not high, and the platform for the combination of production, education and research is relatively immature.
With the emergence of the intelligent era, the sensor industry is at a rare historical opportunity. Seizing this historical opportunity, sensors will usher in a new height of development. For China's sensor industry, it bears a great responsibility and faces major challenges.
To this end, China needs to find breakthroughs in the following aspects: improving sensor accuracy, improving small-low-cost mass production capacity, multi-material composite technology, battery technology and wireless passive sensors, packaging and testing equipment and systems, processing equipment and consumables localization, etc.
At the same time, the establishment of a large ecosystem of intelligent sensor industry requires not only devices, but also testing, processing and other links. Through a strong industrial ecosystem, we will improve the level of China's smart sensor industry.
5 Conclusion
On the one hand, the research direction of smart sensors is to explore new materials, new principles, and new technologies to improve the performance of the sensor itself, and on the other hand, with the integration of sensor technology and standard CMOS process, miniaturization, multi-functionalization, and intelligence will be the inevitable trend of future development.
China should seize the historical opportunity brought by the intelligent era to the development of the sensor industry, comprehensively improve the basic research and industrialization level of intelligent sensors, and provide strong technical support for the arrival of the intelligent era.