Introduction:Now shopping peak is no longer the biggest challenge, where will the next generation of technological innovation appear?
Twelve years after its birth, shuang-11 continued to write the answer paper and left the questionnaire: when the peak shopping is no longer the biggest challenge, where will the next generation of technological innovation appear?
“Good innovation comes from high-pressure scenarios, and” double 11 “creates high-pressure scenarios one by one, so that the most cutting-edge technology and the most cutting-edge business model can have a high-speed collision.” Last night, Cheng Li, chief technology officer of Alibaba group, said in his speech at the closing stage of the double 11 that the technical challenges of the double 11 will enter a new historical stage.
This year’s double 11 history has the longest span. With COVID-19 background, there are more than 2000 industrial belts, 300 thousand foreign trade factories and more than 30 thousand overseas brands. When there are opportunities to push 4000 million new products to 800 million or more consumers, every key link has the possibility of being born with the new generation of technology.
Picture: Alibaba Group CTO Cheng Li’s speech on the night of double 11
Compared with Cheng Li, an engineer who has been involved in the whole process for 12 years, people can’t immediately perceive the full value of technology engineering. When many projects are first designed, they are to save a few minutes and a few yuan for consumers and businesses. In terms of social costs, they need to flatten a peak in time or space. Finally, they return to how to make every order stable and efficient reach.
As Cheng Li said, this year’s double 11 has brought about a lot of innovations in new product development, manufacturing, user access, supply chain, logistics and distribution. And their ultimate direction is to make “business elements online and digital, on this basis, to form a truly digital business network”.
Because the time window is the longest, let’s forecast the new things in the future from a longer industrial chain
C2m technology will replace “e-commerce”, just as cloud computing will replace traditional it
C2m (customer direct connection manufacturing) may still be a term with unclear definition, and there is no consensus on “what c2m technology includes”. However, just like cloud computing companies, replacing it companies, this double 11 makes the future c2m technology cluster clearer.
Alibaba has at least three different technology modules to try to get through c2m in depth.
- The aiot technology and multimodal interaction module behind tmall spirit have affected the intelligent design and development of many home appliance and home furnishing manufacturers.
It has also led to the rapid growth of some small household appliances industrial belts and foreign trade factories.
The overall openness of technology and design resources is still the highest in the whole industry.
- Laiping Zhizao represents the industrial digital technology from the terminal to the local, in which the 3D / AR front-end rendering model, the design software of the home decoration designer and the production database of the building materials factory are connected.
The 100000 model rooms of double 11 are just the beginning, and 20000 stores will enter the whole house customization business in the next half year.
- The intelligent factory and flexible production line of rhinoceros Zhizao with the support of data intelligence technology, the opportunity that rhinoceros Zhizao brings to industrial belt factories is not only the measurement of small samples, but also the rich opportunities under big data.
Home decoration, home appliances, home furnishings and clothing are growing significantly. There is a strong trend of manufacturing flexibility and consumer personality in these industries. The market scale is huge enough. For example, Alibaba has a very strong technical end problem for localized distribution and assembly.
With Alibaba cloud’s accumulation in several basic industries in the industrial brain field, it may not take long for Taobao tmall to hatch a c2m technology cluster or intelligent manufacturing platform from an e-commerce platform.
Digital supply chain becomes the breakthrough of large-scale enterprise digitization
If the cloud is the resource on the dispatching line, the supply chain is corresponding to the offline resource of the enterprise, which is undoubtedly extremely important. Therefore, in his speech, Cheng Li repeatedly quoted a project of Alibaba called dchain. Through the intelligent prediction technology, more than 1.1 billion goods have been allocated to the nearest logistics warehouse to potential consumers in advance before the double 11; some retail enterprises directly save the storage logistics cost and deliver goods directly from the factory; even on the eve of the live broadcast room of the anchor, they can predict where tens of millions of orders may be distributed.
Picture: Cheng Li mentioned the shared service ability of the society in his speech
At present, this system should be accessible to enterprises through rookie digital warehousing system and Alibaba cloud’s new retail platform products. It can be imagined that in the very sinking commodity management such as Taobao special edition or retail link, this ability will greatly improve the supply efficiency of manual management in the past. But is that all?
- The supply plan of an enterprise is the core of the whole business resource decision-making. Compared with the customer relationship problem solved in the past, it was considered that the inventory was not digitized because of the low frequency of inventory in and out.
In fact, it is probably because there are not enough data and algorithms that can really optimize the supply.
- Few institutions’ digital supply chain can gather Alibaba’s multi-dimensional links.
In addition, it has deeply connected the order messages that affect the supply chain with more than 140 service providers in almost all fields;
Once large enterprises can refer to it, it will drive a large number of upstream and downstream ecological enterprises.
- This field does not need new software, but needs new data processing and algorithm framework. For example, in Alibaba business services and data warehouse products, there are a large number of stream batch integrated architecture.
The experience accumulated by double 11 seems to be getting closer to productization.
In addition, there is the new product platform of Taobao tmall, which directly helps the brand to develop early technology platform. For some need to use explosives to determine the enterprise plan of the industry, will be very critical.
The understanding of consumers and industry may give birth to a new intelligent computing framework
At the Alibaba double 11 technology communication meeting on November 3, the important role of cognitive intelligence engine in commercial AI system was explained in detail for the first time. Then, on November 9, the team was invited to release graphscope, an open source one-stop graph computing engine, at a top-level conference sponsored by the China Association for science and technology, the Chinese Academy of Sciences and the Chinese Academy of engineering.
This open source framework hasn’t published many details, but from the introduction of the project team, it comes directly from the intelligent scenario of hundreds of billions of calls per day in double 11
- In Alibaba’s scenario, because there are conceptual maps of industry, region and economic behavior, the demand for computing is different from the general deep learning scenario.
With this ability of traversing graph, intelligent technology with higher cognitive level can be produced for different industries.
- Graph computing performance has improved 10 times in recent years, but there are still three large-scale application challenges
First, problems in related fields are complex, computing modes are diverse, and a large number of solutions are fragmented;
Second, it is difficult to learn, and the threshold for non professional programmers is very high;
Third, the amount of cross domain data is large, and the computational efficiency is still low.
- There will be a collaborative relationship between graphscope and tensorflow in the system, that is, deep learning and traditional graph model will be more closely combined to achieve better results.
Figure: application scenarios of graph computing introduced by Alibaba cognitive intelligence team
The external observation of platform companies is often very different in granularity. Especially for Alibaba technology, everyone touches different parts and has different evaluation. Alibaba seems to be the largest company that is most willing to show its customers the latest R & D output because it has the richest industry: some are in the sky, some are growing up in the ground, some are exploring the architecture concept of Zhongtai, and some cloud habitat conference has continuously thrown out such explosive news as Dharma Institute and pingtouge.
If you go back to one of the company’s strategies: cloud computing and big data, under the catalysis of double 11 over the years, Alibaba’s self-developed stories of cloud computing and database have finally realized this kind of element level technological innovation. Cloud computing in China keeps the first place in both scale and growth, and database in China has surpassed Oracle in the first place; it not only supports the double 11, but also is the infrastructure for many national applications.
Different from the general closed and centralized monopoly platform, the development of high technology must be open and altruistic. From the increasingly abundant double 11 orders, it can be seen that adhering to large-scale open interface and maintaining high-frequency social collaboration is the most sustainable innovation in itself.
This article is the original content of Alibaba cloud and cannot be reproduced without permission.