On June 15, cloud-based healthcare services provider Huami (NYSE: HMI) unveiled its latest chip for wearable devices, Huangshan No. 2, its new sensor BioTracker™ 2, along with an Artificial Intelligence (AI) algorithm for health big data at the the AI Innovation Conference in Hefei, Anhui Province. Huami CEO Huang Wang also announced the launch of a new institute, Huami AI Research.
The Huangshan No. 2 chip is the upgraded version of Huangshan No. 1, which was launched in 2018. The new chip uses the advanced RISC-V structure, the IoT era’s most promising design, which boasts the fastest processing speeds and efficient energy usage. Huangshan No. 2 is faster than its predecessor due to improved local biological data computing technology. For example, it can identify atrial fibrillation 6 times faster than Huangshan No. 1 and about 25 times faster than other chips. It is estimated that the Huangshan No. 2 chip will go into mass production during Q4 this year and that new chip-powered wearable products will debut during the first half of 2021.
AI data engines
In a 2019 letter, Huang Wang mentioned that those keen to make use of data must make sensors themselves. At the conference, Huami showed it has followed through, as it launched the BioTracker™ 2 PPG sensor, which enables Huami’s five latest biological data engines. BioTracker™ 2 is the most accurate human biology sensor Huami has developed so far. With the great AI computing capacity that Huangshan No. 2 offers, BioTracker™ 2 may become the world’s leading wearable sensor. The set of five new biological data engines has brought Huami’s current data analysis to a whole new level.
RealBeatsTM, a biological data engine launched by Huami in 2019, is used to monitor the wearer’s cardiovascular health and identify abnormal signs through heart rate data analysis and ECG (Electrocardiograph) readings. The newly launched RealBeatsTM 2 has improved on the original by eliminating signal noise disruptions from sports and more effectively and quickly monitoring atrial fibrillation.
Another Huami biodata engine, OxygenBeatsTM, monitors blood oxygen saturation, an important vital sign that has gained prominence due to the COVID-19 pandemic. The device marked an important milestone recently, as its readings deviate only 1.67% from those of other blood oxygen monitoring tools available in the market. Partnering with Dr. Zhong Nanshan’s research team, Huami will provide COVID-19 patients follow-up management using OxygenBeatsTM technology to monitor blood oxygen saturation. New smartwatches with OxygenBeatsTM will be available by 2020 Q3.
Studies on sleep data have always been a focal point of Huami’s research. The new SomnusCareTM AI engine makes use of Huami’s large healthcare data pool and will help users better understand their sleep patterns and nap behaviors. Sleep apnea syndrome is a hidden killer, but by pairing OxygenBeatsTM and SomnusCareTM, Huami will help users identify sleep apnea and remind them to seek professional medical advice.
Heart rate, blood oxygen saturation, and sleep pattern monitoring are the main pillars of Huami’s healthcare infrastructure. However, Huami has also set up an epidemic early alert forecast model with its health data and other external factors. With the additional data engines, the model will make epidemic forecasting an essential part of Huami’s healthcare infrastructure.
On top of the vital sign monitoring applications, fitness is another important segment of Huami’s business. At the conference, the company introduced ExerSenseTM and huami-PAI™, two biological data engines that aid fitness activities. ExerSenseTM can automatically identify when wearers are exercising, while Huami-PAITM helps users set reasonable fitness targets based on their age, gender, resting heart rate, and other biological information.
Huami AI Research
Huami’s own chips, algorithms, and sensors are driving the rollout of a comprehensive healthcare management infrastructure. Huami announced that it is now shifting its focus to AI development. The institute, Huami AI Research, will drive innovation in healthcare big data in the wearable device industry by working with the world’s leading scientists and researchers. Huami Vice President Wang Kongqiao is the director of the institute and will be working with Professor Ramesh Jain, a well-regarded AI scientist from the University of California, Irvine. The institute’s recently-formed research committee is chaired by Wu Feng from the University of Science and Technology of China and consists of 7 leading researchers specializing in AI, sensors, and medical and other studies. Huami announced the institute will undertake a strategic partnership with voice recognition startup Unisound to collaborate on smart voice assistant technology.
As Huami pivots to AI research, the company’s development and enhanced research capacity have laid a solid foundation to continue building the healthcare ecosystem in the long run. The institute’s first three labs cover a wide range of research interests, from practical clinical research and fitness training applications to advanced human-machine interactions. It will be exciting to see how Huami further improves its wearable devices and expands its AI research horizons.
(Top photo from Huami; photo features Huang Wang, Huami CEO)