It is the top leadership’s vision for a new Chinese economy in the age of AI.
China aims to mass-produce neural-network processing chips, robots will make accomplishing daily tasks easier for disabled people, and machine learning will help radiologists read x-ray scans. In addition, China hopes AI will make manufacturing more eco-friendly: the goal laid out in the document is to increase the energy efficiency of the manufacturing sector 10 percent by 2020.
The effort to boost AI in Chinese manufacturing is why Andrew Ng, who helped start the Google Brain project and previously served as the head of Baidu’s AI efforts, announced Landing.ai, a company aimed at helping businesses in the manufacturing sector transform themselves into AI companies. Ng picked manufacturing because it has a number of problems that machine learning techniques can help with, and AI has the potential to radically transform the industry.
Action goals
Through the implementation of the four key tasks, it strives to achieve a major breakthrough in a series of landmark artificial intelligence products by 2020, form an international competitive advantage in several key areas, deepen the integration of artificial intelligence and real economy, and further optimize the industrial development environment.
– The scale-up of the key products of artificial intelligence, the substantial increase of the technology level of the intelligent network car, the large-scale application of intelligent service robots, the strong global competitiveness of products such as smart drone, the expansion of clinical application of medical image diagnosis system, Video image recognition, intelligent voice, intelligent translation and other products have reached the international advanced level.
– The core competence of artificial intelligence has been significantly enhanced. Intelligent sensor technology products have achieved breakthroughs in design, foundry and packaging and testing technologies to international standards. The production of neural network chips and large-scale application in key areas have led to the initial availability of the open source development platform Support the rapid development of industry capacity.
–Development of smart manufacturing, complex environment identification, and new artificial intelligence technologies such as human-computer interaction have accelerated the application of integrated applications in key technical equipment. The application of new models such as intelligent production, large-scale personalized customization and predictive maintenance has been significantly improved. The level of intelligence in key industries has been significantly improved.
–An artificial intelligence industry support system is basically established, with a certain scale of high-quality annotation data resource base, the establishment and opening up of standard test data set, the establishment of artificial intelligence standard system, test evaluation system and security assurance system framework, the establishment of intelligent network infrastructure Gradual formation of the system, the industrial development environment more perfect.
Second, cultivate smart products
Driven by market demand, the Group will actively cultivate artificial intelligence innovative products and services, promote the industrialization of artificial intelligence technology, and promote the integrated application of smart products in the fields of industry, healthcare, transportation, agriculture, finance, logistics, education, culture and tourism. Develop intelligent control products, speed up the breakthrough of key technologies, develop and apply a batch of intelligent devices featuring complex environment perception, intelligent human-computer interaction, flexible and precise control, and real-time collaboration among groups to meet the requirements of high availability, high reliability and safety, Enhance the device’s ability to handle complex, unexpected and extreme conditions
Promote the popularization of intelligent hardware, deepen the application of artificial intelligence technology in fields such as smart home, health management, mobile intelligent terminal and vehicle-mounted products, enrich the intelligent functions of end products and promote the upgrade of information consumption.
Focus on the following areas to achieve the first breakthrough:
(A) intelligent network of cars. Support vehicle intelligent computing platform architecture, automotive smart chips, autonomous driving operating systems, vehicle intelligence algorithms and other key technologies, product development, building software, hardware, algorithms integrated vehicle intelligent platform. By 2020, a smart, intelligent, real-time smart car networking platform will be established to create platform-related standards that will support highly autonomous driving (HA level).
(B) intelligent service robot. Support the development of key technologies such as intelligent interaction, intelligent operation and multi-machine collaboration to enhance the intelligent level of domestic service robots such as cleaning, elder care, rehabilitation, disability and children’s education, and promote public service robots such as inspection and navigation and fire rescue robots Innovative applications. Development of three-dimensional imaging positioning, intelligent precision safety control, human-computer interface and other key technologies to support the development of surgical robot operating system to promote the use of surgical robots in clinical medicine. By 2020, breakthroughs will be made in key technologies such as environment awareness, natural interaction, autonomous learning and human-computer collaboration of intelligent service robots. The intelligent home service robots and intelligent public service robots will achieve mass production and application, medical rehabilitation, helping the elderly and helping the disabled, and fire and disaster relief. Robot prototype production, complete technical and functional verification, to achieve more than 20 application demonstration.
(C) smart drones. Support intelligent obstacle avoidance, automatic cruise, autonomous flight for complex environment, group work and other key technologies research and development and application to promote a new generation of communications and positioning and navigation technology in the UAV data transmission, link control, monitoring and management applications, Development of intelligent flight control system, high-integration-specific chips and other key components. By 2020, the intelligent consumer UAV 3-axis mechanical stabilization unit achieves a precision of 0.005 degrees, achieving 360-degree omnidirectional perception avoidance and realizing automatic and intelligent forced avoidance of air traffic control areas.
(D) medical imaging diagnosis system. We will promote the standardization and standardization of medical image data collection and support the research and development of medical imaging aids in the field of typical diseases such as brain, lung, eye, bone, cardiovascular, and breast, and accelerate the commercialization and application of medical imaging aided diagnosis systems. By 2020, the most advanced multimodal medical imaging diagnostic system in our country will detect more than 95% of the above typical diseases, the false negative rate is less than 1% and the false positive rate is less than 5%.
(E) video image identification system. Support biotech, video comprehension, cross-media fusion and other technological innovations, and develop typical applications such as human-animal syndrome syndication, video surveillance, image search and video summarization, and expand application in key areas such as security and finance. By 2020, the effective detection rate of face recognition in complex dynamic scenes exceeds 97% and the correct recognition rate exceeds 90%, which supports the recognition of face features in different regions.
(F) intelligent voice interactive system. Support the innovation and application of the new generation of speech recognition framework, colloquial speech recognition, personalized speech recognition, intelligent conversation, audio and video integration, speech synthesis and other technologies, and carry out popularization and application in such key areas as smart manufacturing and smart home. By 2020, the average accuracy rate of Chinese speech recognition in multi-scenarios will reach 96%, the recognition rate of 5 meters in far field exceeds 92%, and the accuracy of user dialogue intension recognition exceeds 90%.
(Seven) intelligent translation system. Promote the application of high-precision intelligent translation system, and use machine learning techniques to enhance the accuracy and practicability around typical scenarios such as multilingual translation and simultaneous interpretation. By 2020, there has been a clear breakthrough in multilingual intelligent mutual translation. The accuracy of translation of products in Chinese-English translation and English-Chinese translation exceeds 85%, and the accuracy rate of intelligent translation between ethnic minority languages and Chinese is significantly improved.
(H) smart home products. Support intelligent sensor, Internet of Things, machine learning and other technologies in the application of smart home products to enhance the intelligent level, practicality and safety of home appliances, smart network equipment, water and electricity meters and other products, the development of intelligent security, smart furniture, smart Lighting, smart ware and other products, the construction of a number of smart home test evaluation, demonstration projects and promotion. By 2020, the categories of smart home products will be significantly enriched, the penetration rate of smart TV market will reach over 90%, and the intelligent product level of security products will significantly increase.
Third, break the core foundation
Accelerate the research and development and application of high-precision, low-cost smart sensors, break through the neural network chip and supporting tools for cloud training and terminal applications, and support the research and development of artificial intelligence development frameworks, algorithm libraries and toolsets, and support the construction of open source and open platforms. The layout of intelligent software designed for artificial intelligence applications, tamping the hardware and software base for the development of artificial intelligence industry. Focus on the following areas to achieve the first breakthrough:
(A) smart sensor. Support the research and development of key technologies such as miniaturization and reliability design, precision manufacturing, integrated development tools and embedded algorithms, and support the research, development and application of smart sensors based on new requirements, new materials, new processes and new principles. Development of new bio, gas, pressure, flow, inertial, distance, image, acoustic and other smart sensors with promising market prospects and promotion of technological innovations in materials such as piezoelectric materials, magnetic materials, infrared radiation materials and metal oxides, (MEMS) and complementary metal-oxide-semiconductor (CMOS) integration technologies to develop and develop smart sensors based on new principles such as magnetic induction, ultrasound, non-visible light and biochemistry for new application scenarios, Accuracy, high reliability, low power consumption, low cost. By 2020, the performance of piezoelectric sensors, magnetic sensors, infrared sensors and gas sensors will be greatly improved. Acoustical sensors with signal-to-noise ratio of 70dB and acoustic overload point of 135dB will be mass-produced. The absolute accuracy is within 100Pa and the noise level is within 0.6Pa Of the pressure sensor to achieve commercial, weak magnetic field resolution 1pT magnetic sensor mass production. In the simulation, design, MEMS technology, packaging and personalized testing technology to achieve the international advanced level, with mobile wearable, the Internet, automotive electronics and other key areas of system design capabilities.
(B) neural network chip. Development of high performance, scalability, and low power cloud NE chips for machine learning training applications, development of low power, high performance terminal neural network chips for machine learning computations, development and application of neural network chips for terminal applications Supporting the compiler, driver software, development environment and other industrial support tools. By 2020, the breakthrough has been made in the technology of the neural network chip, and the cloud neural network chip with the performance of 128TFLOPS (16-bit floating point) and the energy efficiency ratio of more than 1TFLOPS / w has been introduced. The energy efficiency ratio exceeds 1T OPS / w Benchmark) terminal neural network chip to support one or several mainstream neural network algorithms such as convolutional neural network (CNN), recurrent neural network (RNN), long and short term memory network (LSTM); in intelligent terminals, automatic driving, intelligent Security, smart home and other key areas to achieve large-scale neural network chip business.
(C) open source open platform. For the common technologies such as machine learning, pattern recognition, intelligent semantic understanding, and other key industries such as autopilot, it supports development of cloud-based training and terminal development frameworks, algorithms libraries and toolsets, supports open source development platforms, open technology networks and The construction of open source community encourages the construction of an open computing service platform that meets the needs of complex training and encourages key leading enterprises to build a new industrial ecology based on the open source and open technologies for software, hardware, data and application collaboration. By 2020, the open source development platform for cloud training supports large-scale distributed clusters, various hardware platforms, various algorithms, and lightweight, modular and reliable open source development platforms for terminal implementation.
Fourth, deepen the development of intelligent manufacturing
In-depth implementation of intelligent manufacturing, to encourage a new generation of artificial intelligence technology in all aspects of exploration and application of the industry to support key areas of algorithm breakthroughs and application innovation, system upgrade manufacturing equipment, manufacturing processes, industrial applications, the level of intelligence. Focus on the following areas take the lead in achieving a breakthrough:
(A) intelligent manufacturing of key technologies and equipment. Enhance the self-testing, self-tuning, self-adapting, self-organizing and intelligent level of high-end CNC machine tools and industrial robots, and improve the machining accuracy and product quality of additive manufacturing equipment by using artificial intelligence technology. Optimize intelligent sensors and decentralized control system DCS), programmable logic controller (PLC), data acquisition system (SCADA), high performance and high reliability embedded control system and other control equipment in complex work environment, improve the perception, cognition and control ability, improve the digital non-contact precision measurement, Online non-destructive testing systems and other intelligent detection equipment measurement accuracy and efficiency, enhance the flexibility of assembly equipment. Enhance the intelligent level of logistics equipment such as high-speed sorters, multi-layer shuttle cars and high-density storage shuttles to realize accurate, flexible and efficient material distribution and unmanned intelligent warehousing.
By 2020, the intelligent level of high-end CNC machine tools will be further enhanced. A new generation of industrial robots with human-machine coordination, natural interaction and autonomous learning will be mass-produced and applied. The forming efficiency of the additive manufacturing equipment is more than 450cm3 / h and the continuous working time is more than 240h; realize intelligent sensor and control equipment integration in the fields of machine tools, robots, petrochemicals, rail transit and other fields; the accuracy of industrial field visual recognition of intelligent detection and assembly equipment reaches 90%, the measurement accuracy and speed meet the actual production needs; Develop more than 10 intelligent logistics and warehousing equipment.
(B) a new model of intelligent manufacturing. Encourage discrete manufacturing enterprises to network production equipment, based on intelligence, application of machine learning technology analysis and processing of field data, equipment online diagnosis, real-time control of product quality and other functions. Encourage process-oriented manufacturing enterprises to build the whole process, intelligent production management and security systems, to achieve continuous production, intelligent production safety management. To create a network of collaborative manufacturing platform to enhance man-machine collaboration under the guidance of artificial intelligence and collaboration between enterprises R & D design and production capacity. Develop customized service platform to improve the depth of learning and analysis of the characteristics of user needs, optimize the product’s modular design capabilities and personalized portfolio. Set up a control and automatic diagnosis system based on standardized information collection, accelerate the training and optimization of the fault prediction model and the user habit information model, and improve the life cycle analysis ability of products and core accessories.
Brian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive technology and trends including Space, Robotics, Artificial Intelligence, Medicine, Anti-aging Biotechnology, and Nanotechnology.
Known for identifying cutting edge technologies, he is currently a Co-Founder of a startup and fundraiser for high potential early-stage companies. He is the Head of Research for Allocations for deep technology investments and an Angel Investor at Space Angels.
A frequent speaker at corporations, he has been a TEDx speaker, a Singularity University speaker and guest at numerous interviews for radio and podcasts. He is open to public speaking and advising engagements.