Character Interview | Focus on the industrialization of edge AI and seek "intelligent change" with "integration"

Recently, Mr. Bao Zhiwei, Director of Marginal AI System and Server at Advantech Technology, and Jun Chen, Senior Product Manager of Advantech (China) Industrial Internet of Things Working Group, were interviewed by reporters together, explaining how Advantech focuses on marginal AI industrialization in the era of artificial intelligence and pursues career intelligence with the integration of industrial skills.

From face recognition to behavior recognition, from license plate recognition to enabling transportation, from packaging testing to intelligent warehousing. Nowadays, artificial intelligence skills are ushering in a wave of implementation in many production and life scenarios, profoundly changing the present and shaping the future. Advantech, as a global enterprise with key influences in the automation industry, embedded computers, and the Internet of Things, is taking advantage of the wave of intelligent development to catch up.

Character Interview | Focus on the industrialization of edge AI and seek "intelligent change" with "integration"

　　Empowering Careers Advantech digs deep into the use value of AI segments

Proven by vivid examples, marginal AI is constantly improving power and increasing value for all walks of life. Bao Zhiwei showed that the marginal AI industry is rapidly entering the "industrialization" stage, and its value and potential are being paid attention to by more and more enterprises.

Some time ago, many AI-based electric vehicle elevator monitoring and alarm systems presented in the profession illustrate this point. With the help of mature AI skills, the use of visual structured analysis to complete real-time monitoring and prevention of dangerous behaviors or situations is in line with the continuous release of AI skills to make profits.

According to the actual needs, Advantech has launched an elevator AI safety monitoring system through cooperation with Xindun Technology, and Chen Jun, senior product manager of this project, said, "Based on computer vision and artificial intelligence skills, the system can identify events that endanger passenger safety, and trigger echo functions such as warning prompts and emergency braking, and then reduce the occurrence of accidents." The elevator AI safety monitoring system is equipped with a marginal AI inference module built on Advantech's Jetson platform, which is compact and wide-temperature designed, which is easy to install in messy and demanding use scenarios such as elevators, and has low power consumption, flexible I/O expansion, and support for deep learning models, which is an aspirational choice for the use of marginal computing embedded AI.

Bao Zhiwei said that in the face of the needs of production and life, the elevator AI safety monitoring system is only a microcosm of Advantech's current marginal AI empowerment in thousands of industries. From facial recognition and behavior analysis to unmanned driving, vehicle-road collaboration, robot control, etc., Advantech has taken advantage of the new opportunities of the integration of AI and the real economy to explore the use value of occupational segments, and continuously bring a wider range of perception, cognition, and intelligence innovations to various occupations, so that the professional use of AI can be integrated into every scene around people.

Easier said than done, Advantech considers the value of AI in many ways

Opportunities often coexist with challenges, and the integration and use of marginal AI with thousands of industries is more rigorous, and the problems and challenges are more obvious.

Bao Zhiwei believes that although AI skills have given birth to many new products, new uses, and new forms, the innovative breakthroughs and industrial implementation of artificial intelligence are still important topics at this stage, and industrialization still has a long way to go.

Just as electricity is changing society, AI is also indispensable for "generators", "high-voltage power grids", and power infrastructure suitable for different occupations if it has the ability to satisfy the large-scale use of the whole society.

"It's not just a simple algorithm evolution, but the landing and use of AI industry requires enterprises to have the ability to store and integrate common sense in professional subdivisions." Bao Zhiwei said that taking traffic use as an example, license plates and vehicle identification are only a small part of the implementation of intelligent transportation, and enterprises also need to have a systematic in-depth understanding of the traffic facilities and traffic laws and regulations in the country or a certain region, so as to build a targeted overall solution plan.

The storage of common sense in the early stage is indispensable, and the implementation of the overall solution plan requires the full help of ecological collaboration partners. Bao Zhiwei emphasized that at that time, the flat organizational structure was replacing the previous chain structure, and in the implementation of some professional AI, it is no longer appropriate for a single enterprise to go it alone, and only by building an ecological partner circle can AI be truly empowered.

Bao Zhiwei took the implementation of the intelligent warehousing plan as an example, he thought that from the purchase, code scanning, transmission, delivery, replenishment and other warehousing operations, a single enterprise could not be satisfied with the skill development and algorithm follow-up of all the use links, and the relevant enterprises in the demand and intelligent warehousing ecosystem collaborated with each other to complete the project landing and use. At the same time, Bao Zhiwei pointed out that AI industrialization is a kaleidoscopic and chaotic project, and the needs are deeply adapted to industrial use scenarios, and the needs are adapted to the environmental requirements of use scenarios, such as highways and other intelligent transportation scenarios have high requirements for the network environment, but in reality this requirement cannot be satisfied in remote areas.

Combining differentiated advantages, Advantech accelerates the implementation of software and hardware skills

In the more chaotic use environment at that time, leading enterprises that understand affairs and advanced enterprises that understand skills are discussing and running in together to establish a final plan that meets industrial needs.

Advantech has always been committed to skill development and investment as an accelerator to promote career intelligence innovation.

In the field of hardware, whether it is in the selection of processors, the secondary development of chips, and the selection of accelerated modules, Advantech is not soft on investment. Through collaboration with Intel, NVIDIA, and Huawei, the AI acceleration module was rapidly iterated and deployed, and the value was maximized through the combination of different hardware products under the condition of comprehensive consideration of all factors.

In terms of software, Advantech focuses on selecting independent software developers with the same advantages in the professional field to collaborate, on the one hand, to complete the skill barriers that can be built together, and on the other hand, to meet the customized requirements of more professional users. Of course, this is also related to Advantech's initiative of "specialization in the art industry", which only selects areas with its own strengths for in-depth research.

At the same time, Advantech has cooperated with independent software developers in a certain region, and when the business is mature, with the help of Advantech's advantages in global business expansion, it can complete the pan-regionalization of software development and use in the professional field, bringing value to the collaboration partners and the development of their own maximum extension skills.

With the continuous implementation of more subdivided use scenarios, in addition to the early construction, the importance of later operation protection of occupational intelligence is increasing day by day. Bao Zhiwei introduced that Advantech has worked deeply with software developers in cooperation with NVIDIA, and imported the software developed together into Advantech's DeviceOn system platform to complete remote collaborative management and protection. Advantech also collaborates with world-renowned data and device security companies to complete the comprehensive protection of system data and information resources, helping end users manage and protect the system more easily.

Create a win-win situation Advantech builds a marginal AI mesh value ecosystem

Advantech has been shaping itself as a collaborative partner of system integrators in the process of enabling industrialization and innovation, and has pursued collaboration to create industrial development.

Today, in the marginal AI industry, Advantech's own positioning has not changed, but the value chain of the profession is quietly transforming. Bao Zhiwei pointed out that in the previous chain architecture, the relationship between equipment vendors and integrators, transaction service providers and end users was an upstream and downstream relationship, and there were few cross-transaction relationships with each other. However, in the construction of some large-scale projects, end users often need to cooperate with multiple parties to complete their customized and personalized needs. In particular, with the continuous improvement of transaction collaboration, the chain architecture is no longer out of date, and the network value ecosystem is gradually forming.

Advantech has strengthened collaboration with independent software developers in the business field by cooperating with core enterprises in the upstream chip side, and continuously extending business with system integrators, gradually building a marginal AI mesh value ecosystem with itself as the core, and playing a critical role in connecting the previous and the next in the process of serving end users. "With the continuous use of AI subdivisions at that time, there are also many grassroots start-ups with certain skill advantages in these fields." Bao Zhiwei particularly emphasized that although these companies have skill advantages, their AI implementation ability may lack engineering ability. Advantech has completed the process of "hen with chicks" through a variety of collaborative methods, leading new AI companies in terms of transaction form and implementation capabilities, so as to help them achieve sustainable development.

Character Interview | Focus on the industrialization of edge AI and seek "intelligent change" with "integration"

Sections viewed

Contact

working hours