After a year and a half of "studying ChatGPT", SF Express launched the "Fengzhi" logistics decision-making model. Giants seize the beach, who can seize the opportunity to land?

Source/National Business Daily

Original title/After a year and a half of "studying ChatGPT", SF Express launched the "Fengzhi" logistics decision-making model, and the giants seized the beach, who can seize the landing opportunity?

More than a year after Wang Wei, president of SF Express, publicly stated that "SF has begun to study ChatGPT", SF Express has passed on new news on the large model.

On August 19, the reporter of "Daily Economic News" learned from SF that SF Technology launched its self-developed "Fengzhi" logistics decision-making model on August 18, which aims to apply the large model technology to the decision-making fields such as intelligent analysis, sales forecasting, transportation route optimization and packaging optimization of the logistics supply chain.

As early as last year, when ChatGPT was born and sparked global discussions, Wang Wei publicly expressed enough interest in this technology and a sense of crisis. However, compared with other companies' explorations in the field of logistics models, SF Express is still belated.

Since last year, JD.com and Cainiao have released large models on the logistics industry side. Jingdong Logistics disclosed that in the first half of this year, Jingdong Logistics applied the logistics model to Jingdong's "618" guarantee work for the first time, and as of the end of June, the logistics model has been called more than 30 million times, empowering hundreds of thousands of front-line employees. In March this year, Zhang Hao, CTO of Cargo Lala, publicly revealed that the self-developed freight worry-free large model has been applied in multiple scenarios, with an accuracy rate of more than 90%.

Now, with the addition of SF, the melee of logistics models has escalated again, and who will seize the opportunity in the competition of logistics companies in the era of large models?

01

Take the first step in the layout of the large model

SF Express launched a logistics decision-making model

"Reality is always cruel, to tell everyone the truth, 'Dick Online' (SF's 30th anniversary speech) was not written by me, but by a large model." In the first half of last year, Wang Wei shared a screenshot of the conversation.

"In the process of rapid development of science and technology, SF Express cannot be lazy, and needs to use new things and new tools to not fall behind." Wang Wei said on the internal platform, and bluntly said that "in this era, we are all eaten by the tigers of technology or innovation in minutes, not by our peers."

Now SF has finally launched its own large-scale model products. The reporter of "Daily Economic News" learned that in order to solve the problem of the landing of large models in supply chain operation scenarios, SF Technology integrates the interactive advantages of large language models and the professional depth of traditional small models to build a supply chain agent. Based on SF Technology's Fengzhi Cloud ecosystem, the agent can accurately inform customers of the cause of problems when customer sales fluctuate, provide managers with decision-making basis, and help them choose appropriate response strategies.

SF Technology told the reporter of "Daily Economic News" that the "Fengzhi" logistics decision-making model has built a multi-level and multi-channel demand forecasting model based on the ability of multi-modal large model to achieve more accurate prediction results. Not only that, the change in the calculation method also reduces the number of model requirements and resource consumption, and improves the resource efficiency.

In one case, for example, the server resource requirements were greatly reduced, the runtime efficiency was increased by 120 times, and the prediction accuracy was increased by 5%. This improvement is very difficult to achieve in the traditional model, and it also marks a major technological breakthrough in the field of supply chain demand forecasting of the "Fengzhi" logistics decision-making model.

However, compared with the logistics company that released large-scale model products as early as last year, SF Express is not too early.

"In recent years, general artificial intelligence, especially large language model technology, has made significant progress, profoundly changing the ecology of many industries. However, in the complex and highly specialized field of logistics supply chain, its application potential has not yet been fully tapped and released. Tang Kai, vice president of SF Technology, said.

Tang Kai also mentioned that supply chain operation is a highly professional and rigorous field, and although the current large-scale model technology shows great potential, it limits its in-depth application in this field to a certain extent in terms of accurate calculation and avoiding "illusions".

Tang Kai further said that it is particularly critical that the results of supply chain operation decision-making and analysis must be traceable, traceable and verifiable, that is, the so-called "white box" characteristics, which is fundamentally different from the current "black box" output method of large models. This core difference has become a major challenge for the widespread promotion of large models in the supply chain field, hindering the full use of their capabilities.

It is worth mentioning that on the same day (August 18) that SF Technology launched the "Fengzhi" logistics decision-making model, SF Technology also reached ecological cooperation with a number of large-scale model start-ups, including 010000 things, and SF's further actions and layout in the field of large-scale models have aroused expectations in the industry.

02

Giants Arrive:

The large model has been explored and implemented in multiple logistics scenarios

Who can get ahead of the curve?

The industry believes that in many industrial scenarios, road freight has the characteristics of perfect system, many links, complex data, rich scenarios, etc., and is one of the main fields for the application of large models. The in-depth application of large-scale model technology is expected to bring great changes to the operational efficiency, service experience and business innovation of the whole logistics chain.

Therefore, since last year, rookies and JD Logistics have already taken action to release large-scale model products for the field of logistics supply chain. In June last year, Cainiao Supply Chain released a digital supply chain product "Tianji π" based on large models, which improved quality and efficiency in the fields of sales forecasting, replenishment planning and inventory health through rookie algorithm + generative AI assisted decision-making based on large models, and effectively promoted the logistics supply chain to enter the era of large models.

On July 13 last year, at the 2023 JD Global Science and Technology Explorer Conference, JD officially launched the Yanxi large model, which supports language, speech, vision, multi-modality, etc., integrates 70% general data and 30% of the native data of the digital intelligence supply chain, and is committed to going deep into the logistics industry scene and solving real pain points.

Baidu Maps also announced at the end of September last year that based on Baidu's large model technical capabilities, combined with the characteristics of the logistics industry scenario, it officially launched the beta version of the logistics large model, taking the lead in the application of logistics address analysis and logistics scheduling decision-making. In October last year, Fuyou Truck and Tencent held a signing ceremony to reach a comprehensive strategic cooperation around the creation of the first digital freight model.

In March this year, the self-developed freight worry-free large model of cargo Lala focuses on scenario-based and lightweight, and is positioned as a "logistics expert around you". At present, the accuracy rate of its freight factual Q&A is more than 90%. In the same month, the first alliance in the logistics industry focusing on the research and practice of large-scale model application was established, which was jointly established at the 2024 Digital Intelligence Logistics Summit by Alibaba Cloud, Cainiao, AutoNavi Map, COSCO Shipping, China Eastern Airlines Logistics, YTO Express, Shentong Express, Zhongtong Express, Deppon Express, G7 Yiliu, Overground Railway, and the Institute of Intelligent Transportation of Zhejiang University.

Now, the launch of SF's self-developed large model also means that all the logistics giants are present, and a more fierce war has officially started in the field of logistics large models.

However, Chen Guanling, a technical partner of Fuyou Trucks, once told the "Daily Economic News" reporter that the current logistics model is still in a very early stage, in the vertical industry of logistics, because the products are not creative products, the accuracy of the output of the large model is very high, and the accuracy of the current large model is not enough, the randomness is relatively strong, and more algorithm improvements are needed, as well as a large number of data and scene enrichment, to solve this problem. This is also the biggest challenge facing the current logistics model.

The consensus in the industry is that 2024 is the year when generative AI is combined with the industry, and it is the year when it can really be implemented. Nowadays, with more and more heavyweight guests in the logistics industry entering the game, who can seize the opportunity in the logistics field has also become the biggest attraction and suspense in the future.

This article is from Xinzhi self-media and does not represent the views and positions of Business Xinzhi.If there is any suspicion of infringement, please contact the administrator of the Business News Platform.Contact: system@shangyexinzhi.com