Analysis of the Industry Chain of the Most Intelligent Service Robot: Paying Close Attention to Three Trends

At present, although the popularity of intelligent robots is the general trend, the players in the industry still have to face many uncertainties and death tests.

Industry Chain Analysis of the Most Intelligent Service Robot: Three Trends Need Close Attention

The breakthrough of AI, the decrease of the cost of core components and the emergence of "pioneer" products have led to the rise of intelligent service robots in recent two years. For a time, voice interaction, dialogue and question-and-answer, face recognition, environmental awareness, autonomous positioning and navigation have almost become the standard of intelligent robot products.

In the household field, intelligent robots based on speakers, desk lamps, pets and IP images emerge in endlessly, and begin to enter the family as early education, companionship and intelligent housekeeper. Cleaning robots such as sweeping, mopping and window cleaning also realize barrier navigation and route planning functions, and their performance is constantly improved, and they are widely used in the household field. 。

In the commercial field, intelligent robots began to appear in different places such as banks, hotels, shopping malls, hospitals and schools, playing the role of customer service, welcome guests, shopping guides, assistants and assistants. The UAV and autopilot cars included in the field of large robots are also improving in terms of intelligence, which is expected to reshape the logistics and transportation in the future. Business model and competition pattern of traditional industries such as transportation and travel.

However, due to the immaturity of technology and engineering, the high expectations of consumers, the lack of habits, and the time required to penetrate into traditional industries, the current market of intelligent service robots only sees capital and enterprise efforts, and the popularity of demand-side applications. There are also divergent opinions about the outbreak of the service robots market in the industry.

So, how is the intelligent service robot industry going? What problems still exist? What factors will determine the real outbreak and popularization? This article will help you sort out the current development status of intelligent service robots from the dimensions of industry chain, industry structure, key technologies and industry applications, and answer the above questions for you.

Note: Robots can be divided into industrial robots and service robots according to industry applications, and programmable robots, adaptive robots and intelligent robots (perception, recognition, interaction, judgment, reasoning, planning and learning) according to the degree of intelligence. This paper focuses on intelligent service robots, including home/personal intelligent service robots, business intelligent service robots and so on. Although UAVs and driverless cars are also included, this article will not discuss them for the time being. The relevant industry analysis will be presented to you in the future.

Map of Intelligent Service Robot Industry: Three Levels and Two Kinds of Players

The industry chain of intelligent service robots covers core components such as chips and sensors, AI technologies such as image, voice and semantics, robot operating system and application of upper products.

Industry Chain Analysis of the Most Intelligent Service Robot: Three Trends Need Close Attention

Figure: Map of Intelligent Service Robot Industry

At present, players in the field of intelligent service robots can be roughly divided into two categories: technology companies and product companies.

Technological companies include companies providing AI chips, operating systems, voice, semantics, images, perception and other underlying and core technologies. They represent enterprises such as Horizon, Cambrian Science and Technology, Turing Robot, Intelligence, Cloud Zhisheng, Speech, Jettison Huasheng, ROOBO, Triangle Veterinary Science and Technology, Watergate Amy, Obi Zhongguang, Think. Lan Science and Technology, etc.

Product company refers to the company that develops products in the subdividing industry or application field. It represents the company with Rokid, ROOBO, Kovos, Conrad, etc. in the personal/household field, Tianzhihang, Baihuiweikang, Pangolin, etc. in the commercial field, as well as the excellent and compulsory choice across the consumer and commercial fields. Of course, there are also innovations in Xinjiang, Baidu and dominant technology to make UAVs and UAVs.

Although some companies not only export technology, but also do consumer-grade products (such as ROOBO, going out to ask), and some product companies rely on independent research and development of their core technology, from the perspective of the industry as a whole, technology and products are two completely different tracks, with their own unique judgment dimension.

In technology companies, hardware and core parts manufacturers take upgrading technology and reducing costs as their main tasks, while AI technology companies take algorithms and data as their core competitiveness. For product companies, demand positioning is the core, and product experience, brand, marketing and ecology are important barriers. Therefore, in judging the value and future direction of different companies, we should start from the company's track and its own positioning.

Software and Hardware Technology of Intelligent Robot Supported by Chip, OS and AI

To judge the value of technology companies, we must first understand the technical background, problems and difficulties in the relevant fields.

Intelligent service robots include three main technologies: chip, operating system and AI technology. Among them, AI technology includes perception, interaction and motion control technology. It can be said that each technology is in a different stage of development, with its own problems and challenges.

chip

At present, most intelligent service robots use general-purpose chips, including CPU, GPU and FPGA. Compared with CPU, GPU has better performance in image operation. Although FPGA does not have general computing function, it is not only efficient in simple repetitive operation, but also low in power consumption. It is also widely used in the industry.

In addition to general chips, in recent years, Yingweida, Chinese Academy of Sciences and other enterprises and research institutions have also developed a special AI chip based on deep neural network, which is expected to be widely used in intelligent robots, intelligent driving and other fields. AI Chips Provided in China

Please read the Chinese version for details.