Consumer robots will once again spur the advent of artificial intelligence bubbles

In the past month, the big news in the robot industry has been continually being seen. The well-known robot company led by Tencent announced that it has received a $82 million Series C financing, which has set a record of US$600 million in a single round of financing for Shangtang Technology. The company’s latest valuation has reached US$5 billion, making it the highest global valuation. AI Ventures; Boston Power Company's new version of the robot dog SpotMini, up and down the stairs, will take the initiative to avoid obstacles, you can open the door yourself. Compared with the girl who was two years ago, Atlas is now more flexible, able to run and jump over obstacles. The two robots unlocked the anti-sky skills are amazing and burst into the network. . In addition, Boston Dynamics revealed that the SpotMini robot is in the early stages of production and plans to enter commercial production next year. All these signs indicate that the consumer robot market, which has entered a cooling-off period in 2017, seems to be gradually warming up.

Robots are divided into the top-end uses, including military robots, industrial robots, commercial robots, and consumer robots. This article explores the 2C-end consumer robot. In theory, the consumer robot can have a home scene, a portable scene, an office scene, a car scene, and the like according to the scene. But at present, there are no signs of development outside the home scene, and home intelligence robots can basically represent 2C-end consumer robots.

With the artificial intelligence industry entering the landing period, the technical development of audio-visual semantic understanding, natural language processing, neural networks, and emotion recognition algorithms has enabled consumer-grade robot products to gradually enter the consumer's field of vision. However, the crazy chase of capital, the function of the product, and the endorsement of the Internet giant may not make the industry, especially the humanoid robot, family housekeeping robot and other sub-fields burst into a burst. On the contrary, the artificial intelligence industry bubble will be triggered by the consumer robot industry.

In 2014, it became the "first year of robots", and it was very hot. The situation in which companies entered the robot industry in different ways was almost crazy. According to incomplete statistics, as of the end of 2014, more than 70 listed companies have acquired or invested in robotics and intelligent automation projects, and the number of Chinese robot-related companies has exceeded 4,000. By 2015, this situation has reached its peak, and then it has been declining since 2016, and capital has gone cold.

According to the statistics of the 2017 Sino-US Artificial Intelligence Ventures Status and Trend Research Report released by Tencent Research Institute & IT Orange, domestic intelligent robots and drone-related technology startups are the most popular, ranking first in the echelon. However, the influx of capital has caused the homogenization of most of the service robot products on the market. After entering 2017, many rushed incomers have gradually lost their voices, and the consumer robot market has gradually entered a cooling-off period. At present, smart drones, restaurant robots, virtual assistants, intelligent hardware and other fields and industries have begun to show signs of decline. According to preliminary estimates, the total number of closed enterprises in China and the United States has exceeded 50. The virtual assistant technology similar to Amazon Alexa and Google Assistant is not fully mature, and the smart robot consumer market for installing applications has not been opened.

Capital is profitable, product homogeneity is serious

As the development of the robotics industry has been promoted to a national strategy, the favorable policies and high prospects of high subsidies have led to numerous capitals entering the industry. However, the robot industry is different from the Internet and mobile phone industries. It has a large investment, slow effect and high risk. It can't be iterated in a few months, only a little progress. The robot is a comprehensive and diversified platform. If the robot is to replace the manual work, it needs a lot of data calculation and command analysis, and the sensors, measurement strength and other technologies need to be coordinated to realize. Due to the bottleneck of the supporting technology, the robots made in China are more like toys, and the fields that can be fully applied are narrow.

From the perspective of capital, its ultimate goal is to obtain commercial value, which makes it "short-sighted." Counting the products of the consumer robot startup company established in China in 2011, except for a few clearly positioned as programming robots and fighting entertainment robots, most of the products have entertainment, early education, programming, housekeeping and many other functions, making the home intelligent robot function on the market. Stacking, getting together and competing.

For many manufacturers, the goal is to be a desktop robot and access virtual assistant technology. The appearance is nothing more than a mechanical combination of flat, wheel, arm and so on. In order to impress users, there will be a huge amount of educational content, video calls and so on at the software level. In the eyes of consumers, such products are dazzling and varied, and can achieve many so-called "high-tech" functions, ranging from a few hundred dollars to several thousand dollars; in the eyes of developers, these are all It is a selection stack for mature solutions. Just like building blocks, from the shape, mobility, to the screen, the camera, to see who can spell out a burst. These products are mixed, the essence of which is intelligent early education machines, or intelligent toys, but wearing a robotic hat. Therefore, a large number of "robots" currently on the market cannot be called robots, and more are to make ends meet and defraud profits. The threshold of real robot products is very high, so it is imperative to implement and standardize.

If you only analyze from the media reports, you can hardly establish the correct understanding of the robot industry, because the network is full of all kinds of remarks that advocate and sing. "The era of consumer robots has arrived: a trillion-dollar market," "consumer-level robots are melee, education and companionship or first to enter the family"... The fundamental reason behind this is that everyone is not talking about a concept: the robot industry does It is the size of the trillion market, but the family consumer robot is not.

As early as 2014, some experts predicted that "China will become the largest market in the global robotics and intelligent equipment industry, and the industry driven by robots will open a new blue ocean of trillions." According to the robot report released by the international consulting firm BCG in the same year, it is also 100 billion yuan to convert the entire robot market into RMB by 2025. For the consumer robotics sector, it was only $2.5 billion in 2015. BCG specifically stated that the consumer robots here refer to robots such as entertainment, cleaning, education, etc. The largest share of the robots is the sweeping robot.

Recently, GfK also released the latest report that China's consumer-grade robots will exceed 10 billion yuan in retail sales in 2018. The most mainstream products are sweeping robots, smart speakers, civilian drones, and intelligent service robots. Category, the best sales are still sweeping robots, accounting for nearly 5 billion yuan in retail sales, and the share of intelligent service robots, only 1 billion yuan.

China's consumer-grade robot retail scale (in 2018, smart service robots are only 1 billion yuan). Source: Zhi things

Therefore, from the beginning to the end, home intelligence service robots are a real market. It's just a little smaller. Its global market size is about $1 billion. Compared with the toy market, it is only a one percent reduction. A large number of entrepreneurs poured into this small mouth, let the market heat up sharply. In just one year before and after, the Blue Sea smashed into the Red Sea, and many start-ups ushered in a wave of closure.

At present, there are many companies that focus on the development of humanoid robots and family housekeeping robots, and they attract huge capital injections, but the field should be the most cautious for investors and practitioners. Although Tencent is in the forefront, the entire track investment is excellent, Leju, Ganker, but it does not help, because capital can not give birth to an industry that does not exist.

In 2017, the price war between Xiaomi and Tmall speakers enabled the voice assistant to accelerate into thousands of households. With the launch of the low-cost version of the small Alpha egg of the University of Science and Technology, the family robot officially entered the price of 100 yuan to kill, so that the thousand yuan market is not high. Within two years, more mature family children's equipment will go to market. When Amazon, Xiaomi, and Ali are based on speakers, and differentiated products into this market segment, who will buy more than a thousand dollars, redundant models of robots? After all, in addition to walking, other functions can be achieved by the speaker + screen / camera series.

Robot products should not be separated from the actual application of the floor. Japan's new humanoid robot, its main function is to take a straw, hand cup, support people to sit on a wheelchair, the price of synthetic renminbi is about 1.3 million, the service target is a nursing home or a disabled person in a nursing home. Although the function of taking the straw seems to be really powerful, how important it is for the disabled is a question that everyone deserves. As for the simulation robot like Sophia, it has been spurred by the AI ​​god Yann Le Cun as a complete scam, except that it can make gimmicks and nothing else.

I know that many people are interested in this question: What is the significance of studying humanoid robots? What is the future?

Among them, the highest praise: For scientific research institutions, humanoid robots are not the "dedication research direction" in the field of robotics, but a sample showing the results of research and development of robots; for industrial companies, they are more inclined to make very human-like Robots, because customers will like it.

This question and answer confirms from some point the human curiosity about the human-like machine with intelligence and the question of whether the human form of the robot is redundant. The head of a company that manufactures robotic performers in the UK responded to interviews and encountered the same questions. Humanoid robots will be used as entertainment and commercial venues, but will not undertake the main tasks. After all, most artificial intelligence can be physically free. jobs.

3. Returning to the user's underlying demand for consumer robots can kill a bloody road

In the face of a small market and massive competition, only by forgetting the robot's ontology and starting from the actual needs of users can we highlight the encirclement.

Aside from the chaos of the product, the consumer's underlying demand has remained unchanged for decades. Whether it is a narrow emotional companion robot or a generalized sweeping robot, the underlying needs are nothing more than three: education, entertainment and efficiency; the corresponding underlying values ​​are to get education, happy entertainment and improve efficiency.

Educational value refers to the value of early childhood education tools.

In the educational robot market, the combination of robotics and programming education, software and hardware interaction has become the mainstream product form of programming education. In this field, LEGO's EV3 is a pioneer. Microsoft, Google, Apple and other large companies have regarded programming education as a part of their long-term ecosystem. The domestic makeblcok is also a model of value.

With the development of the Internet and artificial intelligence, the entire toy market has also undergone rapid and intelligent iteration. In 2017, Sony launched the innovative toy toio, which made people shine. Nintendo developed a paper case kit labo for switch, which is a burst of friends. In 2017, Lego sold tired in the world, but ushered in the country. It is a continuous double-digit growth...Intelligent toys, at least the best place for entertainment when the robot dreams of going down.

The value of the tool is to let the machine take off the sci-fi coat and return to the essence of the machine's tool attributes.

From the generation of sci-fi family robots to the generation of smart speakers, the small fish from the mermaid to the screen speakers at home, are the embodiment of the red sea separation and value return. Abandoning the illusory coat and returning to the essence of the tool is also the essence of returning value.

Home service robots want to get the most practical application scenarios, from the perspective of smart home, use AI to enable intelligent service robots to complete intelligent home control (door and window management, lighting management, access control, energy management, etc.), home appliance control (TV, Audio, rice cooker, air conditioner, water heater and other controls) and security protection (anti-theft monitoring, gas leakage, emergency alarm, etc.), thus becoming one of the control terminals of smart homes, is a reliable choice for returning value.

In my opinion, the main reason why human beings spare no effort in robot research is not to entertain and consumerize it. The first requirement is to realize the value of robot tools and let it do something or meaningless instead of human beings. Or dangerous, or auxiliary things, such as security robots, aerial work robots, underwater robots, etc. Therefore, for the moment, the market for consumer robots that face education and entertainment alone will not be large, and the overall industry standards are different, product quality is uneven, and competition is fierce. For capitalists and practitioners, Risky, investment needs to be cautious. Only by returning to the real needs of users and working hard to make products can it be swallowed up by the overflowing bubble in this round of AI waves.