"There is a big barrier to the Internet of Things in China. Enterprises including Huawei , Alibaba , Baidu have their own set of so-called intelligent platforms, and these are among the leading ecosystems. There is no connection,” Sun Ji, President of Yitian Intelligence, told China Business News reporter.
In the smart home market, many companies have entered the market in the form of single-product smart terminals and launched products with a wide variety of functions to meet the diverse needs of customers. However, the compatibility of equipment between companies is poor and the brand ecosystem is fragmented. increasing day by day. Data provided by the "2021 Global AIot Developer Ecosystem White Paper" shows that on average, each household uses more than 40 electrical devices, involving more than 10 brands. As far as the smart Internet of Things is concerned, the reason for the fragmentation between equipment and brand ecology is that it is difficult to interconnect between various platforms.
This is not a technical problem.
“Mainly needs to be negotiated at the commercial level, such as opening up protocols and opening up platforms.” Sun Ji said that the company operates its own middle platform because it hopes to access more platforms. He believes that all application-oriented companies in China will I hope that major platforms can connect with each other to gain more compatibility.
Difficulties in upgrading single product intelligence
Yitian is a company that specializes in integrated stoves. It has been researching intelligent equipment since 2014. After 2 years of research and development, it launched the first generation of intelligent products in 2016. This product basically realizes intelligent interconnection, implants a large number of recipes, and realizes voice and entertainment interactive functions.
"At that time, there were already some ready-made module solutions on the market, such as the intelligent central control system . We just did more functional linkage and customized development on the ready-made solutions." Sun Ji said that the product had not yet been realized at that time. The Internet of Things in the entire kitchen scene is just the most basic form of interconnection.
Since then, Yitian has superimposed more functional modules on the integrated stove products. From the initial implantation of WiFi modules and voice recognition modules, it has expanded to value-added services such as image recognition, visual intelligent final screens, entertainment, and shopping.
For kitchenware companies, the investment in intelligence is huge.
takes the configuration of R&D personnel as an example. It requires software engineers, hardware engineers, interaction designers , professional testers, and structural engineers. Investments at other levels, including platform development costs, APP development costs, software and hardware development costs, procurement costs, etc., are all factors that must be considered to achieve intelligence.
"Integrated stoves integrate hoods, stoves, steamers, oven lighting systems, etc. The functions are very complex. When it is to be intelligent, the logic behind it will become more complicated, and all the logic must be run through. To fix the bugs, a lot of testing needs to be done behind the scenes The workload of testing, verifying and correcting errors is huge. "Sun Ji said that in the kitchen scene, during the cooking process, the integrated stove and hood will produce noise, which will interfere with voice control. , this is different from the living room, which is a quiet environment. How to achieve the best voice control effect in a noisy environment is a problem that requires a large number of scientific researchers to invest in solving.
Currently, Yitian’s key work is focusing on how to realize the interconnection of some functions of peripheral products in the kitchen scene, and in the future, we will explore how to integrate the intelligence of the whole house. For example, it can be interconnected with curtains, LED lights, etc. in the living room.
The development path of the home Internet of Things can be roughly divided into three stages: single-product intelligence (various brand manufacturers launch a series of smart single products), whole-house intelligence (from focusing on control needs and single device networking to evolving to whole-house intelligent experience and applications) organic integration of scenes), full-scenario intelligence (from residential as a carrier to user-oriented, extending from inside to outside the house, and multi-device interconnection control to be scene-oriented).
Many companies are stuck in the process of upgrading single-product intelligence to whole-house intelligence.
The current pain point of AloT implementation is the high degree of fragmentation of downstream application scenarios and needs, resulting in heterogeneous IoT terminals and diversified network communication methods and platforms.CICC’s research report believes that if the problem of fragmentation is not solved, it will be difficult to realize the interconnection and interoperability of different equipment manufacturers and products in the AIoT ecosystem, and the original intention of reducing costs and increasing efficiency will not be achieved.
“99% are small companies”
Businesses don’t always know what consumers really want from the Internet of Things. In the process of upgrading to product , it is not uncommon for things to become intelligent for the sake of intelligence.
Take electric ovens as an example. Seeing the rise of the Internet of Things concept, some manufacturers have migrated several cooking mode buttons on traditional ovens to mobile phones, claiming that the oven has achieved intelligent transformation. This actually increases the cost of use for users. Users may still need to turn over regularly and visually observe the completion of the food in the oven, and the operating devices have increased from an oven to an oven and a mobile phone. Pseudo-smart products such as
are often priced higher than traditional products. Not only do they fail to bring the convenience of smart IoT to users, but they also increase the learning cost and burden caused by cumbersome operating procedures.
"Limited by technological conditions, the functional appeal of electric ovens appeared at that time, or the product manager thought of such a solution, which just represents the product manager's thinking about the function at that time." Zhang Wei, general manager of Tmall Elf AIoT, said that the existing Intelligent products can be divided into two stages: one is connection. After the data is online, the App can control the device, which can lead to many scenarios, such as air-conditioning energy consumption issues. The App can control and save power and achieve energy consumption management. This is difficult to achieve with traditional remote controls. Second, after the connection is realized, users can remotely operate the device, manage data, and achieve more functional interactions. The R&D and innovation of
product functions are never achieved overnight.
"In the process of making the first-generation oven intelligent, limited by the feasibility of technology, it was necessary to wake up the voice, turn on the oven, cook dishes, adjust the temperature, and adjust the time. Each action required the user to send a voice command, which was too much. It's cumbersome and cumbersome to interact with. It is very unfriendly to user interaction and does not improve efficiency. "Sun Ji said that now through an intelligent recipe association, you only need to tell Tmall Genie to make steamed fish, and the product will start directly because of the temperature behind it. , time and other logic are all set.
Sun Ji said that when the company designs and considers intelligence, the first concept it adheres to is true pragmatism; the second is to simplify the complex.
In recent years, IoT products and demonstration applications have emerged one after another. The number and output value of related enterprises have increased year by year, and certain results have been achieved. However, there are still some difficult problems that need to be solved urgently. At present, it is still a small and scattered industry.
"99% of China's Internet of Things companies are small companies. China's consumer Internet of Things has not outgrown large companies, not entirely because the underlying technology is not well developed." Senior member of IEEE, Professor of Hohai University Han Guangjie believes that the main reasons include: first, the development of the Internet of Things industry chain is uneven, and there is a lack of communication operators similar to the three major domestic ones. Doing business with leading IoT enterprises such as Huawei and ZTE ; secondly, it is difficult for upstream and downstream enterprises in the relevant industry chain to collaborate, and there is a lack of unified government functional departments to coordinate various IoT enterprises, formulate unified standards, and work together ; Third, the market positioning is vague and the profit model is uncertain. The Internet of Things currently has problems such as market fragmentation and limited scale, scale and standardization have not yet been formed, and the application cost of Internet of Things technology is too high.
Many companies have launched IoT platforms and used private cloud or public cloud services to manage the interconnection of devices. The main participants include operating system providers (such as Huawei, Google , etc.) and application software and Integrated solution providers of hardware services (such as Tuya, Xiaomi , Tmall Elf, etc.), as well as numerous cloud business service providers such as Amazon , Alibaba , etc. For Chinese companies going overseas, their interconnection solutions usually require the help of overseas platforms such as Amazon. For example, the IoT functions of TCL products sold overseas are built on the basis of Amazon cloud technology.
CICC's research report believes that the rapid development of the intelligent Internet of Things industry requires the support of platform enterprises and new operating system software to break fragmented information islands.
Who is the ultimate entrance?
There are different paths for interaction between consumers and smart devices, such as voice and vision. In the early days, voice interaction was the first choice for platform-based enterprises.
Take the smart home field as an example. Tmall Elf, Xiaomi and Baidu initially used smart speakers as their entry points.
Data provided by Xiaomi shows that in the first quarter of 2022, the monthly active users of Xiaoai classmates reached 115 million, a year-on-year increase of 36.2%. Most of them came from users of Xiaomi mobile phone , Xiaoai speaker and other users. Tmall Genie's product matrix for smart speakers is diversified, including basic versions priced at more than 100 yuan and speakers focusing on sound quality priced at more than 1,000 yuan.
The main providers of smart speakers in China are Baidu, Tmall Genie and Xiaomi Group. They opened the way with price subsidies in the early days, and their low prices also satisfied users' early adopter psychology, thus quickly opening up sales. Today, these companies sell tens of millions of speakers every year.
However, in terms of smart speakers alone, most of these companies have not achieved profitability.
Xiaomi told China Business News that smart speakers are only a part of the business ecosystem behind it. Simple smart speaker products can provide users with limited usage scenarios. Relying on the ecosystem behind it, it can further provide users with complete ecological services. On this link, smart speaker products are essential.
took 3 to 4 years from the start of smart speakers to the opening of the ecosystem. It started with controlling the on/off function of the device, and gradually extended to functions that enrich scenes, such as opening curtains and chatting with Tmall Genie. , listen to current news, and set sleep mode at night.
Today, most smart speaker manufacturers have begun to add screens to increase visual interaction on top of voice interaction. Sun Ji said, “Tmall Elf provides a complete set of AIot solutions based on our needs, including the Tmall Elf service provider system and the operating middle office that can develop exclusive apps. Alibaba itself can provide very powerful voice, Image recognition and other technologies. "He believes that if you do it yourself, you need to build your own team and develop many functions. At this stage, the company does not have this capability in the short term.
Currently, more than 200 categories can realize device intelligence through Tmall Genie. Regarding the screening of connection categories, Zhang Wei told China Business News that on the one hand, on the user side, it depends on which devices users are more willing to interact with and what they are more willing to control with voice? For example, controlling lights, air conditioners, and TVs; on the other hand, it is to see which products can be interacted with by voice to create new scenarios, such as massage equipment. Users are blindfolded during the experience and it is inconvenient to control the equipment. If using Voice control is more convenient. Playing music during the massage can also reduce the noise generated by the massage. Is the
visual interaction mode replacing the voice interaction mode? The relevant person in charge of Xiaomi Group told China Business News that the basic function of smart speakers is still voice command execution feedback, and visual interaction is a supplement to voice interaction and can bring more content usage scenarios.
At present, the demand for iterations from smart speaker users is weakening. IDC data shows that in the first half of this year, China’s smart speaker market sales were 14.83 million units, a year-on-year decrease of 27.1%. Xiaomi believes that 3C digital consumption is affected when the economy is weak, and the multiple scenarios of smart speakers are highly overlapped with real estate and other industries, which inhibits the demand for new users; in addition, new products in the industry do not provide obviously differentiated services. There is insufficient driving force for users to switch phones.
If smart speakers cannot become the ultimate entrance to the Internet of Things, do AR devices have this potential?
Previous AR devices have been ignored by ordinary consumers due to their high technical requirements and high prices. Its application is currently mainly concentrated on the enterprise side, such as Microsoft Hololens 2 (industrial version) priced between 30,000 and 40,000. Domestic companies have also begun to launch some new AR products or adapt to new scenarios.For example, Lenovo launched the Morningstar Smart Dog in mid-August, which combines AR technology with a four-legged robot and is used for power inspection of substations ; Nreal also launched two new AR glasses at the end of August.
enterprise-level customers have some practical AR application scenarios and can pay higher prices, which is crucial for the emerging AR technology. Lenovo's Morningstar Smart Dog is currently launched in Hebei, State Grid, and realizes functions such as all-terrain inspection, 3D autonomous navigation, remote collaboration, robotic arm-assisted operation, and automatic charging. Mao Shijie, president of the Shanghai branch of Lenovo Group Research Institute, who was interviewed by reporters, believes that AR/VR has the potential to become an important starting point for industrial intelligent transformation.
Some institutions have given optimistic estimates on the market size of AR. Investment institution ARK predicts that the AR market will expand from less than US$1 billion today to US$130 billion in 2030. Whether this prediction can be realized and whether AR can become a super entrance to the Internet of Things will depend on whether the computing power algorithm improves fast enough and whether enough application scenarios can be discovered.
