Core stuff (public number: aichip001)
Author | ZeR0
Edit | Desert Shadow
As people pay more attention to physical health, sales of consumer electronics in the health sector are rising. To collect various vital signs data, professional sensors are indispensable.
From wearable devices such as smart watches, glasses, and rings to smart home products such as smart speakers, air conditioners, and air purifiers, they have the ability to monitor various health data directly or non-contact by equipping them with many different types of sensors.
In the field of smart health, Infineon has implemented the mass production of its sensors in a variety of consumer health management and monitoring functions through a series of partners. Recently, a number of Infineon Technologies Greater China Power and Sensing Systems (PSS) business units exchanged views with zhidong and other media, sharing some practical applications and successful cases of their sensors in smart health monitoring.
According to Li Guohao, Application Marketing Director of Infineon Technologies Greater China Power & Sensing Systems Division, data from ABI, a municipal research institute, shows that the demand for home Healthcare (health management in home scenarios) products is expected to be about 20 million units in 2022, and may rise to more than 180 million units in the next 10 years.
First, four types of sensors simulate human senses and cover mainstream health monitoring scenarios
Infineon's PSS division has four widely used sensors that simulate the human body's senses of hearing, vision, touch, smell and more, thus simulating the actual environment and sensing it.
The first category is MEMS products, microphones can simulate human hearing, in mobile phones, headphones, smart speakers have a wide range of applications. For different application scenarios, Infineon has a wide range of MEMS products.
The second category is 3D stereoscopic radar sensors. It simulates vision and can be applied in the automotive field to detect obstacles from a distance. Infineon has also worked with major customers to develop a radar for industrial and consumer applications in smart spaces, smart homes, home health monitoring, and more.
The third category is the 3D ToF time-of-flight sensor. It can also simulate vision, detect the distance and depth of the human body or object, and can be applied to the mobile phone to determine whether it needs to be unlocked, and can also be applied to the sweeping robot to detect the distance of obstacles.
The fourth type is the environment, air-related sensors, mainly carbon dioxide sensors, can simulate the sense of smell, used in air conditioning and air purifiers, auxiliary equipment for indoor air conditioning, while through the indoor air quality to detect whether there is a presence in the room.
In addition to medical and health devices, more and more wearable devices are also equipped with various types of sensors for measuring data such as steps, heartbeat, blood glucose, and blood oxygen. The health data collected from the sensor can be uploaded to the cloud through the device in the home scene and combined with the existing online medical services.
Taking the scene of caring for the elderly as an example, the sensor can monitor the breathing, heart rate and other vital signs of the elderly at any time, and can be uploaded to the cloud for relatives and friends in other places to know the health status of the elderly. The sensor can also track through the path to detect unexpected situations such as the fall of the elderly in time.
Second, radar: monitor subtle vital signs, perceive distance tracking gestures
Millimeter-wave radar can simulate the human eye, perceive the surrounding environment, especially for some motion detection, has higher sensitivity and reliability.
It can sense any small movements, including tiny displacements in the chest cavity during breathing, and thus monitor some vital signs. It can also use "presence sensing" to stimulate some device systems to perceive data information such as distance, position, speed, and angle, and after combining with algorithms, it can realize the tracking application of breathing, heartbeat, gestures, etc.
For example, Infineon's partners combine their radar sensors with silicon microphones and AI algorithms to make solutions to collect the body's micro-movements in sleep and snoring sounds, thereby monitoring sleep quality and determining whether sleep apnea is present.
Radar waves can penetrate any surface and shell that does not have a conductive substance, so even if it is installed inside the product housing, its working effect is not affected. At the same time, it is not affected by light, and can operate normally in a lightless, high temperature, foggy, and dusty environment.
A major difficulty of radar is the algorithm, as shown in the figure above, the horizontal axis part is for some of the original data information that radar can extract, and the vertical axis part is the working principle of radar and the algorithm popularized on the market. Based on these algorithms, users can implement the functions they want to do. The algorithm is different, and the MCU or processor required for radar is also different.
Infineon's two millimeter wave radars can basically cover the main application scenarios of radar: Infineon's newly launched BGT60LTR11AIP, which is a cost-effective solution for mainly attacking motion detection, which can use automatic mode for detection; the other program is the BGT60TR13C, which has been on the market for a period of time, suitable for some advanced radar sensing technology, it belongs to ultra-broadband, can use advanced FMCW algorithm, the resolution is very fine, can achieve gesture control, vital signs monitoring, Features such as people tracking.
Specifically, the BGT60LTR11AIP is a millimeter-wave Doppler effect-based radar sensor that integrates a simple baseband and uses integrated antenna packaging technology to help customers solve the complexity of antenna design, RF and radar signal handling.
In automatic mode, the radar can detect people at a distance of 5 meters, a horizontal viewing angle of about 80 degrees, and an average power consumption of less than 5mW. In SPI mode, an external MCU or PC can extract raw radar data for use in custom algorithms that have been developed. In this way, the distance detected can be increased to 10 meters away.
Compared with infrared solutions, millimeter waves are more sensitive and have a lower false alarm rate, which will greatly improve the user experience. At the same time, millimeter waves have the characteristics of penetration and non-conductivity, so they are easy to place in the package of some products.
The chip is small in size, about 3×7 mm, the architecture is one shot and one receive, and the antenna is integrated in the package. In auto mode, it has four quad-state input pins for plug-and-play use by designers.
The BGT60TR13C differs from the previous BGT60LTR11AIP hardware architecture in that it is a one-shot, three-receiver MMIC that also uses integrated antenna packaging technology. Its size is 6×5 mm, the average power consumption when used is about 20 to 30 mW, the use of bandwidth is greater than 5GHz, the resolution can reach about 3 cm.
Because it belongs to the architecture of multiple sends and multiple receipts, it can also extract angle information data, using advanced algorithm development, which can realize gesture recognition, indoor personnel presence perception, heartbeat monitoring and other functions, suitable for smart buildings, smart homes, TVs, notebooks and elderly care and other scenarios.
Third, ToF sensor: good at accurate mapping, better protection of privacy
Similar to radar, ToF is also a 3D sensing. The difference is that radar uses millimeter wave, can't see, and has stronger adaptability; While ToF needs to have a light source, which has higher resolution than radar and relatively large power consumption.
ToF through the way of light-sensitive imaging of objects, mainly used to detect depth information, by calculating the time delay of light sent and received multiplied by the speed of light and then divided by 2, you can calculate the distance between the object and the object.
As long as it involves applications that obtain information such as depth, distance, etc., ToF can help. ToF does not collect color image information like CMOS image sensors, but only depth information, which makes it better protect privacy.
For example, ToF is suitable for face recognition functions, such as face feature point calculation and gesture tracking; it can also be used to map the entire scene when the robot moves, to achieve accurate obstacle avoidance, and to provide more data input to the upper-level path planning algorithm.
At present, Infineon's ToF sensors have been mass-produced and supplied in mobile phones, automobiles, robots and other applications, which can provide customers with complete and mature solutions.
Today's Applications of ToF in smart cockpits are very popular, including driver face recognition, passenger seat adjustment, automatic light on and off, and outside obstacle avoidance monitoring. ToF can also help intelligent parking, determine the obstacle situation within a few meters of the surrounding area, and provide blind spots for ultrasonic, RGB, etc. to detect the surrounding environment.
PMD and Infineon are pioneers in the field of ToF and have 20 years of experience.
As early as 2002, PMD has been engaged in related research. Prior to its formal cooperation with Infineon, PMD focused primarily on industrially oriented applications. In mid-2013, Infineon began working with PMD on the main need to make ToF's imager smaller and lower in power consumption and to apply it to consumer products.
Infineon has a lot of experience in cmOS processes, announcing the first generation of ToF imager, called REAL 3, in 2015 and the second generation in 2016. The product was then mass-produced on Lenovo and Asus phones based on Google's Tango standards, and was applied to a home robot developed by Lighthouse, an American startup.
Its sixth-generation product, launched in 2021, has the ability to detect at a longer distance to better adapt to the metaverse trend of AR applications.
It is reported that there are not many ToFs on the market that can really be mass-produced. While many suppliers can only provide low-resolution ToFs, which are often referred to as "range finders," Infineon can provide high-resolution area array ToFs, which are an imager that provides a more detailed image of an object for upper-layer applications.
Infineon and pmd have more than 400 patents in the field of ToF. In the case of strong light, Infineon's ToF can better avoid the interference of sunlight. Infineon has an SBI patent that removes unwanted light information and allows valid light information to come in, so that it can get better depth information about objects under strong light.
In the smart health scenario, Infineon's ToF has achieved mass production of fall detection functions in some nursing homes and isolation hotels.
Fourth, CO2 sensor: monitor air quality, optimize ventilation control
The PPM reading output by the CO2 sensor can reflect the CO2 (carbon dioxide) concentration value.
People tend to overlook the impact of CO2 on our lives and work. In crowded and poorly ventilated environments, CO2 concentrations rise rapidly. For example, a person in a space of about 4 square meters, the CO2 concentration will rise from 500 PPM to more than 1000 PPM in 45 minutes.
When the outdoor air circulation, the PPM of CO2 is relatively low, but when the CONCENTRATION of CO2 is getting higher and higher, such as more than 1000PPM, people will feel a little sleepy, difficult to concentrate, higher and even make people feel headaches or breathing difficulties. If the concentration reaches more than 2000 PPM, people's cognitive ability will be affected.
Therefore, it is necessary to add the function of measuring CO2, targeted monitoring and control of air quality, which helps to improve people's health, improve the comfort of life and improve work efficiency; at the same time, the ventilation system can be controlled and used as needed, thereby saving energy and reducing the operating cost of the ventilation system.
With MEMS technology and experience, Infineon has successfully developed new CO2 sensors based on photoacoustic spectroscopy to provide more universally applicable SOLUTIONs for MEASURING CO2 concentrations.
As shown in the figure above, there are four main technologies for CO2 sensors, in addition to photoacoustic spectroscopy, there are NDIR, EC and eCO2. Sensors based on photoacoustic spectroscopy and NDIR enable the measurement of true CO2 concentrations independent of other gases.
The advantage of photoacoustic spectral CO2 sensors is that they can achieve the same performance as high-end NIDR CO2 sensors, meeting stringent air quality detection standards while saving more than 75% of space.
Compared to NDIR sensors, photoacoustic spectral CO2 sensors use a highly sensitive MEMS microphone as a detector, reducing device and design complexity. As a result, it can be small in size and low cost, making it ideal for high-volume industrial and consumer applications.
In contrast, EC and eCO2 have weaker accuracy and directivity and are susceptible to humidity.
As can be seen from the figure above, Infineon's photoacoustic spectral carbon dioxide sensor is very small, which can achieve 1/4 of the current typical NDIR sensor. It adopts SMD package, in line with international JEDEC standards, can use reflow soldering technology for leadless surface mounting, and does not need to do later calibration after assembly, which is convenient for large-scale production, which can save time and cost in assembly for customers.
Diagonally above is the chamber, which has an infrared light source, optical filters and MEMS microphones, in which the highly sensitive MEMS microphone is used to detect very subtle pressure changes generated by CO2 molecules in the chamber, and only a small amount of gas is required to make an accurate judgment, so the chamber can be designed accordingly in a small size. In addition, the sensor integrates an MCU that converts the detected CO2 concentration value directly into a PPM value, which is output through three interfaces.
All the key components of Infineon's photoacoustic spectral CARBON DIOXIDE sensors, including infrared light sources, optical filters, MEMS microphones, MCUs, MOSFETs, etc., are in-house products developed by Infineon.
Conclusion: Smart health management is emerging more business opportunities
In general, in the smart health scenario, Infineon's various types of sensors can be combined with their MCUs, Bluetooth, Wi-Fi and data security chips, combined with relevant software and partners' algorithms, to achieve consumer AIoT devices with both sensing, wireless and other functions, making smart health monitoring and management more accurate and effective, more simple and easy.
Depending on the business model required for different application scenarios, Infineon can adopt flexible cooperation methods and expand many types of partners to cover the fragmented IoT market as much as possible and continue to develop new applications. The data collected by various sensors is playing an increasingly important role in health management, and the corresponding business opportunities are also increasing.