Recently, the first session of Sequoia China’s “Innovative Compound Interest” Sequoia Talk Series Forum on Science and Technology was successfully held. 4 senior Sequoia China investors and 8 innovative entrepreneurs shared the innovation process in industrial software, robotics, cloud computing and other fields.
Zhou Kui, Fuxin, Zhang Han, partners of Sequoia China, and Zhai Jia, managing director of Sequoia China, brought wonderful conversations; Qu Kaifeng, founder of Beijing Yundao Zhizao Technology Co., Ltd., Chen Zhaopeng, founder of Siling Robot, and Mechaman Shao Tianlan, founder and CEO of DeRobot, Li Di, CEO of Xiaoice, Deng Yuanbo, CEO of Ruiweishi, Ye Zhengsheng, CEO of Jiuzhang Arithmetic Technology Co., Ltd., Zheng Wenbin, CEO of Cyber Kunlun/Founder of Kunlun Lab, and Li Jingyuan, CEO of KAWO Wonderful sharing.
Taking science and technology as the core theme, from the history of technology and industry, to the development status, and then to the future outlook, investors and innovators deconstruct the development context and jointly explore the innovation driving force behind it.
Here is a summary of the conversation:
Yundao Zhizao Technology Co., Ltd.
Topic: Digital Twins: Industrial Software Requires “Apple Thinking”
Moderator Zhou Kui, partner of Sequoia China, guest Qu Kaifeng, founder of Beijing Yundao Zhizao Technology Co., Ltd.
In the context of a new round of scientific and technological revolution and industrial revolution, the development of industrial software has entered a new process. As the infrastructure of the digital economy, industrial software can be the soul and brain that links various tools, machines, equipment and other industrial bodies, and is a necessary support for the digital transformation of industry and manufacturing. We believe that cloudification, platformization, openness and ecologicalization are important evolution directions of industrial software.
Dr. Qu Kaifeng
Industrial software is hatched from high-end industries. In the fields of power electronics, aerospace, automotive and marine, industrial software has become a very basic design and development tool. At the same time, the transformation and upgrading of China’s manufacturing industry will spawn a large number of localized industrial software to meet the needs of small and medium-sized enterprises for low cost, high efficiency and flexibility.
The integration of Internet technology and industry is driving changes in the form of industrial software products, and industrial software in the Internet era will pay more attention to distribution. From the perspective of CAE simulation, industrial software is transforming from a general-purpose software bottom layer to an engineering module oriented to vertical subdivision scenarios, and more and more industry-oriented industrial software has sprung up. For example, the combination of CAE, CAD and the chip field produces EDA software, and the combination with the construction field produces BIM software. On the other hand, with the help of the industrial Internet “PaaS platform + APP” model, a large number of product-oriented industrial APPs have also begun to emerge, gradually forming a simulation ecosystem.
Topic: Intelligentization of Robots: The Difference Between Robots and Humans
Moderator Fu Xin, partner of Sequoia China, guest Chen Zhaopeng, founder of Siling Robot
In the context of the transformation and upgrading of the manufacturing industry, the industrial robot industry has ushered in opportunities for rapid development. Industrial robotics integrates multi-disciplinary high-tech technologies such as computer, control, machinery, mechanism, information and sensing technology, artificial intelligence, and bionics. It does not simply replace the work of human beings, but integrates the specialties of people and machines. It not only has the ability of people to quickly respond to environmental conditions, analyze and judge, but also have machines that can work continuously for a long time, with high accuracy and resistance to harsh conditions. environment, etc.
Sequoia China has been investing in the field of industrial technology for more than ten years, and we have a deep understanding of the pain points in various fields. Based on the insights accumulated over long-term investment, although many people think that robots are already a field of the red ocean, we believe that in the field of high-end robots, especially high-end robots that solve the pain points of the industry, there is still a blue ocean.
Dr. Chen Zhaopeng
Whether compared to the 80 million cars per year in the automotive industry or the 1.4 billion mobile phone market in the mobile phone industry, robots are still a relatively niche and future market. The robot market capacity is not large so far, and we think the reason is that its intelligence is not high enough.
Siling Robots very much hopes to use its own artificial intelligence technology to enhance the capabilities of robots, so that robots can have wider applications in more fields. So the whole Siling robot is more like an iceberg. The part above the iceberg is actually the robot that everyone sees, but it is more useful to make the robot intelligent and more useful, and that is our part. Software and intelligence capabilities. Robot is a field with a huge blue ocean future, which also means that we will face huge challenges, so Siling Robot hopes to move forward side by side with its peers in the sea of stars and stars of robots.
Theme: Crossing the Chasm: From Robotic Innovation to Sustainable Business Operations
Guest: Shao Tianlan, founder and CEO of Mecamand Robot
The robot industry has developed rapidly in recent years. Robots of different shapes and markets are booming, and many giant companies are also actively participating in it. But the commercial success we see for robotics companies is still rare—there’s a huge gulf between good technology, good products, and successful business operations. The traps that robotics companies often fall into include: product definition does not match market demand, overestimation of current technological capabilities and speed of progress, premature expansion when immature, non-standardized projects leading to ineffective growth, core competencies and business Pattern mismatch, etc.
To solve these problems, companies not only need to have strong technical ability, but also need to be able to calculate accounts, understand the entire value chain and find their own position, but also need to have good strategic determination. The current supply and demand side of the robotics industry is in a period of opportunity. As long as there is better patience, the innovative compound interest can lead the robotics industry to cross the chasm and move towards a bright future.
Theme: Virtual Humans Out of the Circle, Embracing Trillion-Level Interaction Demands
Moderator Zhai Jia, Managing Director of Sequoia China, and guest Li Di, CEO of Xiaoice
Can machines think? This question we are discussing repeatedly now, Turing, the father of artificial intelligence, published an article to discuss it in 1950. The answer to this question synthesizes iterations of AI technology.
By the time XiaoIce was launched in 2014, we saw that machines had become proficient in interacting with humans through deep learning. This has also gone through different intermediate forms such as artificially designed scripts and human interaction, artificial intelligence markup language to customize chat content, etc. By 2020, virtual people and digital people will begin to appear in various application scenarios, and the applied technology will also be more Complex, including 5G, motion capture, 3D modeling, speech recognition semantic understanding, etc. Based on these technical iterations, we have seen that the biological characteristics of virtual human skin and hair texture have become more and more realistic, the interactive ability of virtual human has also been enhanced, and the work that can be undertaken has become more and more.
XiaoIce has definitions for robot IQ, EQ, and Personality, but the so-called EQ of a robot is a romantic term. If we use a technical term, EQ actually refers to the dialogue system of artificial intelligence. level to control the conversation.
In the past, chatbots, dialogue robots, etc. that we often talk about are often based on dialogue systems or task completion system actions. The dialogue scenes with them are more like Q&A. Human users put a Question to it, and it can have an Answer. But in real life, the real value of human interaction is not answering questions, but in the entire communication process, in work, in other daily social processes, we humans can also guide the direction of this communication at the same time, Be able to guide the other party to talk about some new issues.
Emotional intelligence is actually a very rational ability, which is very different from emotionality. So can emotional intelligence fit? If we can pre-train a large amount of human sample data to obtain the AI being in the cold start stage, and then let it continue to interact with the outside world and continue to iterate, it will eventually form his emotional intelligence. A significant difference between emotional intelligence and knowledge is that the ability of emotional intelligence is never-ending and is in a process of continuous iterative optimization.
Topic: Vehicle Virtual Reality Fusion Interface: AR-HUD
Guest: Deng Yuanbo, CEO of Ruiwei Vision
Dr. Deng Yuanbo
As an important part of the smart cockpit, the penetration rate of HUD is rapidly increasing, but there are still many pain points that have not been resolved. At present, in order to match the installation of the HUD, it is necessary to make customized changes to the ordinary windshield, and eliminate the influence of ghosting through the use of a wedge-shaped film. Ruiwei pays attention to the technical innovation of principle, and pioneered the parallel vector (PV) projection technology, which can achieve high-definition and ghost-free projection without any processing on the windshield, and has passed the mass production certification.
This innovation can bring two values to OEMs and the industry: First, it can help OEMs reduce costs significantly; second, HUD can be applied to more models, such as passenger cars, trucks, and even rail transit.
Nine chapters arithmetic
Theme: Cloud computing, opening the second half
The host is Zhang Han, partner of Sequoia China, and the guest is Ye Zhengsheng, CEO of Jiuzhang Arithmetic
The concept of cloud computing first started in 1996, and it has gradually developed into a hot topic in the past ten years. After 30 years of development, cloud computing has changed from a concept to a dynamic industry.
With the rapid development of the digital industry, the cloud computing market has also ushered in an explosive period, becoming a very important business incremental part for many startups and large companies. Among them, there are not only large companies’ deep cultivation in this field, but also some A large number of start-up companies are leading cutting-edge technologies in related fields.
The development of cloud computing has entered a multi-cloud era, and cross-cloud data processing is a new challenge. Cloud-native technology will bring fundamental changes to the technical architecture and business model of basic software such as databases. Today, database vendors are all thinking about cloud computing-based architectures that separate storage and computing to become cloud-native systems, and develop toward serverless to achieve better elastic scaling and pay-as-you-go. With the rapid development of the Internet and the digital age, there will be new technological breakthroughs in the fields of multi-modal data management, real-time data analysis and transmission.
In addition, data security is a major shortcoming in the industry, and new ideas often emerge, but in many areas, security testing has not yet been resolved. I believe that there will definitely be big technological breakthroughs in data security, especially in the field of cloud computing.
Topic: Vulnerability Drives Cloud-Native Security
Guest: Zheng Wenbin, CEO of CyberKunlun
Cloud native is a new field, and it also faces many new security risks. CyberKunlun, on the other hand, starts from vulnerabilities and helps customers solve overall cloud-native security problems. Vulnerability is a process of continuous confrontation, not a simple problem of several vulnerabilities. Therefore, CyberKunlun hopes to become like a power station to continuously provide our knowledge and capabilities about vulnerabilities to customers, or to show it through our products.
In terms of cloud-native cloud security solutions, we have three concepts, and provide customers with emergency solutions from three directions: assessment, confrontation, and defense. The core point of this is to be able to achieve the classification and visibility of visible threats; the second step is to confront vulnerabilities; the third aspect is real-time response to attacks and automatic defense.
Topic: Insight into the current situation, starting from needs, reaching out to pain points, how to create products that users love
Guest: Li Jingyuan, CEO of KWAO
In the increasingly competitive market environment, the use of SaaS tools to assist marketers in their work has become an industry trend. So how to create a product that can really help users and be loved by users? KAWO’s answer is: starting from the underlying usage scenarios, build a product matrix upwards.
KAWO has investigated more than 40 companies in the market and found that there are often barriers to communication between the CMO and the executive team. The two parties have different goals for “digital transformation”. Only by understanding the real pain points of front-line users, whether How small the demand is, is the basic basis for more targeted development and optimization of products. The success of an enterprise’s digital transformation is also the result of a two-way process of internal bottom-up promotion and top-down promotion. This is also the opportunity and direction for us to create products and provide services.
This article is reprinted from: https://www.leiphone.com/category/industrynews/Z3rpIg0DPqVi4pNj.html
This site is for inclusion only, and the copyright belongs to the original author.