广告
加载中

工业世界模型首展;易买工品等三家公司IPO同日折戟丨产业互联网周报

亿邦动力 2026-08-02 16:59
亿邦动力 2026/08/02 16:59

邦小白快读

EN
全文速览

本文是最新一期产业互联网周报,汇总了近期业界动态、投融热点、宏观政策与行业趋势,核心干货如下:

1. 业界核心动态:多家头部企业发布AI相关布局,京东工业推出覆盖全业务流程的“具身智链”,宣布未来五年将采购410万台智能设备;卡奥斯推出支持中文需求生成代码的AI开发平台;万联易达提出“产业智能体互联网”新概念,字节、阿里也完成了ToB办公AI业务的组织调整。

2. 核心数据与趋势:目前中国HTTP请求中机器流量占比达59.2%,已经超过人类流量;大模型发展从拼参数转向落地应用,工业AI从单点试验走向系统原生,数据要素成为AI时代核心基础设施。

3. 投融与政策:三家企业同日IPO折戟,也有企业成功登陆北交所;九部门发文支持科技金融领域数据开发,市场化数据机构迎来新机遇。

本文梳理了产业互联网与AI领域最新动向,能为品牌商把握趋势、布局业务提供参考,核心干货如下:

1. 技术与行业趋势:当前机器流量占比已经超过人类流量,大模型发展重心从拼参数转向实体场景落地,已经深度渗透制造、农业等实体经济领域,工业AI已经从单点测试走向原生系统布局,品牌商可抓住窗口跟进AI落地。

2. 可利用的免费资源:企查查面向创业者免费开放3个月价值2000万元的MCP智能体数据服务,目前已经接入腾讯、阿里两大主流AI生态,品牌商开发自有AI智能体可免费使用该资源。

3. 资本与政策红利:当前产业互联网进入价值重估阶段,资本更看重盈利性,品牌商需优先搭建成熟盈利模型;政策层面九部门支持科技金融数据开发,多地成立持牌供应链金融机构,品牌商可借助新的融资服务缓解资金压力。

本文汇总的行业动态中,和卖家相关的机会、风险、政策干货整理如下:

1. 政策带来的新机遇:九部门发文加强科技金融领域数据开发利用,拥有合规资质的市场化数据机构迎来场景扩容,卖家可借助合规数据服务更好开展客户拓展、风险管控等工作;青岛已经成立首家持牌供应链金融科技公司,可通过应收账款、订单融资等产品为上下游卖家提供资金服务,缓解资金周转压力。

2. 高增长市场机会:全球AI算力集群建设需求释放,相关赛道卖家业绩增速明显,飞速创新上半年净利润预增超六成;跨境领域新兴市场仍有增长空间,头部跨境企业已经在印尼3C配件市场拿到第一的市场份额,卖家可关注这两类赛道机会。

3. 风险提示:当前IPO市场冰火两重天,产业互联网进入价值重估阶段,卖家需要尽快跑通“模型+数据+场景”的商业闭环,提升自身盈利能力,规避估值与竞争风险。

本文的行业动态对工厂推进数字化转型、挖掘商业机会有较多启示,核心干货如下:

1. 数字化转型的低门槛机会:当前工业AI已经从单点试验走向系统原生,多家平台推出低门槛工业AI工具,降低工厂转型成本,比如卡奥斯推出的COSMO Code AI开发平台,支持工厂用中文描述需求,AI自动生成工业应用代码,哪怕工厂没有专业开发团队,也能快速将生产创意转化为智能应用。

2. 订单对接的商业机会:京东工业发布“具身智链”布局,明确宣布未来五年将采购410万台智能设备,做智能设备研发生产的工厂可对接该需求,获取大额稳定订单。万联易达已经推出产业智能体协同平台,能解决AI+产业落地难的问题,工厂可接入该平台提升上下游协同效率。

3. 转型方向启示:AI时代数据要素已经成为核心基础设施,工厂转型要重点绑定数据、大模型与自身生产场景,才能构建新的竞争力。

本文梳理的最新行业动态,能帮助To B服务商把握趋势、挖掘客户需求,核心干货如下:

1. 行业发展整体趋势:当前产业互联网的竞争已经从过去的规模比拼,转向智能体能力与生态协同能力的较量,AI+产业已经从单点试验阶段走向系统原生阶段,客户需求已经从单一AI工具升级为全链路智能系统解决方案,服务商需要及时调整业务方向,布局原生智能系统。

2. 新技术与窗口期机会:当前行业已经涌现出“具身智链”“产业智能体互联网”“自然语言驱动工业开发”等多个新方向,万联易达提出,解决单个智能体落地问题后,当前正是布局产业智能体互联网的关键窗口期,多智能体协同将成为下一个需求增长点。

3. 客户痛点与可对接资源:企业AI转型的核心痛点是开发门槛高,卡奥斯的低门槛开发模式已经验证了需求;同时AI智能体需要商业数据支撑,服务商可对接企查查已经接入主流生态的MCP数据服务,完善自身产品能力。

本文汇总的行业动态,能为产业互联网平台把握市场需求、调整运营策略提供参考,核心干货如下:

1. 市场对平台的核心需求:AI时代下,平台需要构建“模型+数据+场景”的完整商业闭环,同时开放生态连接上下游参与者,头部平台的模式可参考:京东工业同时扮演大额需求方与开放操作系统,既放出410万台智能设备采购需求带动上下游,又开放能力连接硬件与场景,卡奥斯开放AI开发平台吸引开发者丰富生态。

2. 头部平台的最新调整方向可借鉴:字节跳动整合飞书、豆包与火山引擎团队,强化AI企业服务的协同;阿里推出千问办公,将大模型能力深度整合到办公场景,可见整合大模型能力到核心业务场景,是当前平台运营的核心方向。同时数据是核心基础设施,开放数据合作接入合规商业数据是必然趋势。

3. 风险规避提示:当前IPO市场冷热不均,资本更看重盈利性,平台需要尽快跑通盈利闭环,同时重点打造生态协同能力,才能在竞争中占据先机。

本文梳理了2026年7月下旬产业互联网领域的最新动态,总结了行业新特征,对产业研究有较高参考价值,核心干货如下:

1. 技术与产业新动向:技术层面,工业AI已经从单点试验走向系统原生,大模型发展重心从“拼参数”转向“重落地应用”,最新数据显示中国HTTP请求中机器流量占比达59.2%,已经超过人类流量,AI正式深度渗透实体经济各个领域。模式层面,行业涌现出“具身智链”“产业智能体互联网”等新概念,万联易达提出为每个企业配置自主运转的企业经纪人智能体,解决AI落地难的行业痛点。

2. 资本层面新特征:当前IPO市场呈现冰火两重天的态势,产业互联网进入价值重估阶段,资本越来越看重企业的盈利能力,多家未盈利企业IPO折戟,盈利性良好的企业成功登陆资本市场。

3. 竞争格局新变化:数据要素成为AI时代的核心基础设施,商业数据已经成为AI智能体的标配能力,产业互联网竞争已经转向智能体能力与生态协同能力的较量,“模型+数据+场景”的闭环能力成为核心竞争力。

返回默认

声明:快读内容全程由AI生成,请注意甄别信息。如您发现问题,请发送邮件至 run@ebrun.com 。

我是 品牌商 卖家 工厂 服务商 平台商 研究者 帮我再读一遍。

Quick Summary

This article is the latest issue of the Industrial Internet Weekly, which aggregates recent industry developments, investment and financing trends, macro policies and sector trends. Key takeaways are as follows:

1. Core industry developments: Multiple leading companies have announced AI-related strategic layouts. JD Industrial launched "Embodied Intelligence Chain", a full-process solution covering all business workflows, and announced that it will procure 4.1 million smart devices over the next five years. COSMOPlat launched an AI development platform that supports code generation from Chinese prompts. Wanlian Yida proposed the new concept of "Industrial Agent Internet". ByteDance and Alibaba have also completed organizational adjustments to their B2B office AI businesses.

2. Core data and trends: Machine traffic now accounts for 59.2% of all HTTP requests in China, surpassing human traffic for the first time. Large model development has shifted from competing on parameter counts to real-world deployment; industrial AI is moving from standalone pilot tests to system-native integration; and data elements have become the core infrastructure of the AI era.

3. Investment, financing and policy updates: Three companies withdrew their IPO plans on the same day, while another successfully listed on the Beijing Stock Exchange. Nine central government departments released a policy document supporting data development in the science and technology finance sector, bringing new opportunities for market-oriented data institutions.

This article summarizes the latest developments in the industrial internet and AI sectors, offering actionable insights for brands to track trends and plan strategic layouts. Key takeaways are as follows:

1. Technology and industry trends: Machine traffic has now surpassed human traffic, and large model development has shifted its focus from parameter competition to real-world industry deployment, with deep penetration into the real economy including manufacturing and agriculture. Industrial AI has evolved from standalone pilot tests to native system-wide integration. Brands are advised to leverage this window to advance their own AI implementation.

2. Free accessible resources: Qichacha has opened three months of free access to its MCP agent data service, worth 20 million RMB, which is already integrated into the mainstream AI ecosystems of Tencent and Alibaba. Brands building their own AI agents can access this resource for free.

3. Capital and policy dividends: The industrial internet is currently entering a phase of value re-evaluation, and investors are prioritizing profitability. Brands should prioritize building a mature profit model first. On the policy front, nine government departments support data development in science and technology finance, and licensed supply chain finance institutions have been established in multiple regions. Brands can leverage new financing services to ease capital pressure.

This article summarizes the opportunities, risks and policy insights relevant to sellers from the latest industry developments, as follows:

1. New policy-driven opportunities: Nine government departments have released policies to strengthen data development and utilization in science and technology finance, opening new use cases for market-oriented data institutions with compliant qualifications. Sellers can leverage compliant data services to improve customer acquisition and risk management. Qingdao has launched its first licensed supply chain fintech company, which offers capital services to upstream and downstream sellers through products such as accounts receivable financing and order financing, helping ease cash flow pressure.

2. High-growth market opportunities: Global demand for AI computing cluster construction is expanding, driving strong revenue growth for sellers in related tracks: Feisu Innovation projects its net profit will grow more than 60% in the first half of the year. Emerging cross-border markets still offer room for growth, with leading cross-border companies already capturing the top market share in Indonesia's 3C accessories segment. Sellers are advised to watch for opportunities in these two tracks.

3. Risk reminder: The current IPO market is sharply divided, and the industrial internet is in a phase of value re-evaluation. Sellers need to quickly build out a complete "model + data + scenario" business closed loop, improve profitability, and avoid valuation and competitive risks.

The industry updates in this article offer valuable insights for factories advancing digital transformation and tapping into new business opportunities. Key takeaways are as follows:

1. Low-barrier opportunities for digital transformation: Industrial AI has evolved from standalone pilot tests to system-native integration, and multiple platforms have launched low-barrier industrial AI tools that cut transformation costs for factories. For example, COSMOPlat's COSMO Code AI development platform allows factories to describe needs in Chinese, and the AI automatically generates industrial application code. Even factories without professional in-house development teams can quickly turn production ideas into intelligent applications.

2. Business opportunities for order matching: JD Industrial announced its "Embodied Intelligence Chain" strategy and a clear plan to procure 4.1 million smart devices over five years. Factories developing and manufacturing smart devices can connect to this demand to secure large, stable orders. Wanlian Yida has launched an industrial agent collaboration platform that solves the pain point of difficult AI+industry deployment, and factories can access the platform to improve collaboration efficiency across the supply chain.

3. Guidance on transformation direction: Data elements have become core infrastructure in the AI era. To build new competitive advantages, factories need to align their transformation with the combination of data, large models and their own production scenarios.

This article summarizes the latest industry developments to help B2B service providers capture trends and identify customer demand. Key insights are as follows:

1. Overall industry development trend: Competition in the industrial internet has shifted from competing on scale to competing on agent capabilities and ecosystem collaboration. AI+industry has moved from standalone pilot testing to the system-native stage, and customer demand has upgraded from single AI tools to end-to-end intelligent system solutions. Service providers need to adjust their business direction in time and prioritize building native intelligent systems.

2. New technology and window of opportunity: Multiple new directions have emerged in the industry, including "Embodied Intelligence Chain", "Industrial Agent Internet" and natural language-driven industrial development. Wanlian Yida notes that after solving the problem of single-agent deployment, we are now in a critical window for building the Industrial Agent Internet, and multi-agent collaboration will become the next growth area for demand.

3. Customer pain points and connectable resources: The core pain point for enterprises undergoing AI transformation is high development barriers, and COSMOPlat's low-barrier development model has already validated market demand. Meanwhile, AI agents rely on commercial data support. Service providers can connect to Qichacha's MCP data service, which is already integrated into mainstream AI ecosystems, to enhance their own product capabilities.

This article summarizes recent industry developments, offering insights for industrial internet platforms to identify market demand and adjust operational strategies. Key takeaways are as follows:

1. Core market demand for platforms: In the AI era, platforms need to build a complete "model + data + scenario" business closed loop, and open their ecosystems to connect upstream and downstream participants. Leading platforms offer reference models: JD Industrial acts both as a large demand aggregator and an open operating system. It drives the entire ecosystem with its 4.1 million smart device procurement requirement, while opening up its capabilities to connect hardware and real-world scenarios. COSMOPlat opens its AI development platform to attract developers and enrich its ecosystem.

2. Actionable insights from leading platform adjustments: ByteDance has integrated its Feishu, Doubao and Volcano Engine teams to strengthen collaboration for AI enterprise services; Alibaba launched Qianwen Office to deeply integrate large model capabilities into office scenarios. This makes it clear that integrating large model capabilities into core business scenarios is the current core priority for platform operations. In addition, data is core infrastructure, and opening up data cooperation to access compliant commercial data is an inevitable trend.

3. Risk mitigation guidance: The current IPO market is uneven, and investors prioritize profitability. Platforms need to quickly complete their profit closed loop and focus on building ecosystem collaboration capabilities to gain a competitive edge.

This article summarizes the latest developments in the industrial internet sector in late July 2026 and identifies emerging industry characteristics, offering high reference value for industrial research. Key insights are as follows:

1. New technology and industry trends: On the technical side, industrial AI has evolved from standalone pilot tests to system-native integration, and large model development has shifted focus from "competing on parameters" to "prioritizing real-world deployment". Latest data shows that machine traffic accounts for 59.2% of all HTTP requests in China, surpassing human traffic, marking that AI has deeply penetrated all segments of the real economy. On the business model side, new concepts such as "Embodied Intelligence Chain" and "Industrial Agent Internet" have emerged. Wanlian Yida proposes configuring autonomous operating enterprise broker agents for each company to solve the long-standing industry pain point of difficult AI deployment.

2. New capital market characteristics: The current IPO market is sharply divided, and the industrial internet is entering a phase of value re-evaluation. Investors are increasingly prioritizing corporate profitability: multiple unprofitable companies have failed to complete their IPOs, while profitable companies have successfully listed on capital markets.

3. New changes in the competitive landscape: Data elements have become core infrastructure in the AI era, and commercial data is now a standard capability for AI agents. Industrial internet competition has shifted to competition over agent capabilities and ecosystem collaboration, and the closed-loop capability of "model + data + scenario" has become the core competitive advantage.

Disclaimer: The "Quick Summary" content is entirely generated by AI. Please exercise discretion when interpreting the information. For issues or corrections, please email run@ebrun.com .

I am a Brand Seller Factory Service Provider Marketplace Seller Researcher Read it again.

# 业界新事 #

1 京东工业亮相具身智能供应链生态大会,发布“具身智链”布局

7月30日,主题为“智链未来携手共赢”的ESCC具身智能供应链生态大会在北京经济技术开发区举办。大会由北京经济技术开发区机器人和智能制造产业局主办,京东工业、富士康科技集团等协办,汇聚300余家具身智能产业链上下游企业及50余位行业专家。京东工业相关负责人在大会上展示了融合“采、存、标、训、评、仿、测”七个环节的京东“具身智链”布局,搭建覆盖全业务流程的闭环体系。京东同时扮演“超级甲方”——未来五年采购410万台智能设备,以及“行业操作系统”——通过JoyInside连接硬件与场景。

2 卡奥斯亮相Agentic AI Summit,推出AI智能开发平台COSMO Code

7月24日至25日,Agentic AI Summit超级智能体系统架构峰会在深圳举行。卡奥斯开源社区在大会上推出面向工业开发者的AI智能开发平台COSMO Code,以“一句话秒出应用”的低门槛模式,帮助开发者快速将创意转化为工业智能应用。平台支持自然语言驱动编程,用户用中文描述需求,AI自动生成代码、修复Bug、迭代功能。卡奥斯数据智能首席科学家余锦泽博士围绕“本体×大模型×多智能体”介绍了工业互联网场景下的可信落地实践。7月30日,卡奥斯还入选2026年IDC中国AI 50强。

3 万联易达“产业智能体互联网”WAIC全球亮相

2026年7月,万联易达在上海举办的2026世界人工智能大会上正式提出“产业智能体互联网”概念。万联易达集团副总裁杜新凯表示,构建能“自动驾驶”的产业智能体互联网,撕开了AI+产业落地难的现实裂缝。万联易达已于4月推出产业智能体协同应用平台,并运营AI大模型“万联摩尔”。在万联易达构建的产业智能体互联网中,每个企业将拥有代表自身利益、能自主运转的“企业经纪人智能体”。

4 字节跳动启动ToB业务最大架构调整,飞书与豆包整合

7月30日,字节跳动启动面向AI业务的组织调整,加强豆包、飞书、火山引擎在企业生产力场景的产品与服务协同。根据调整方案,飞书产品团队与豆包产品团队将整合,成立新的豆包产品团队,由豆包负责人赵祺负责,飞书负责人谢欣向赵祺汇报。GTM体系方面,飞书GTM团队与火山引擎团队整合,成立新的ToB GTM组织“创造力服务平台”。

5 阿里推出千问办公,钉钉CEO上任46天首款产品

7月27日,阿里推出千问办公产品。这是陈宇森自6月11日接任钉钉CEO之后推出的首款产品,从接任到上线仅46天。目前企业微信已接入超1400万真实企业与组织,每天服务微信用户超7.5亿。

6 人民网发布《中国智能互联网发展报告(2026)》

7月29日,人民网研究院在北京发布智能互联网蓝皮书《中国智能互联网发展报告(2026)》。发布会上,中国工程院院士、紫金山实验室首席科学家刘韵洁援引数据称,中国HTTP请求中机器流量已占59.2%,超过人类流量。报告指出大模型发展重心从“拼参数”转向“重应用”,深入渗透农业、制造、医疗等实体经济领域。

7 中国资源循环集团落子江西,中部首家再生金属现货交易市场投用

7月31日,中国资源循环集团旗下中资环有色金属(江西)有限公司在江西贵溪揭牌成立,这是中国资环在江西的首家专注于再生有色金属的实体企业,其参与运营的中部首家再生金属现货交易市场同日投用。

8 企查查免费开放2000万数据资源,已接入腾讯、阿里主流AI智能体

7月30日消息,企查查面向全国OPC创业者免费开放3个月、价值2000万元的智能体数据平台MCP专项服务。目前,腾讯ima的MCP广场、阿里智能体工作台QoderWork等主流AI产品均已完成对企查查MCP的接入。

# 投融热点 #

1 华泰证券首次覆盖震坤行,给予“买入”评级

7月21日,华泰证券发布震坤行首次覆盖报告,给予“买入”评级,目标价4.30美元。华泰预计震坤行2026—2028年营业收入分别为104.1亿元、118.3亿元及135.8亿元,经调整净利润分别为1.5亿元、3.1亿元及4.9亿元,公司有望于2026年实现年度盈利转正。

2 易买工品港股IPO招股书失效

7月30日,易买工品于2026年1月30日递交的港股招股书满6个月,正式失效。公司是中国领先的数字化FA工厂自动化零部件专业提供商。

3 斯坦德机器人第三次递表港交所

7月27日,斯坦德机器人向港交所递交上市申请书,这是其继2025年6月和2026年1月两次递表失效后的第三次冲刺。公司2026年前4月收入同比增长139.4%,但连年亏损。

4 千岸科技7月29日登陆北交所

千岸科技(920065.BJ)于7月29日在北京证券交易所上市,发行价为24.3元/股。公司2026年上半年预计营收9.4-9.6亿元,归母净利润1.10-1.16亿元。(上海证券报)

5 粤十数智再度递表港交所

7月29日,冷链数智化龙头粤十数智再次向港交所主板递交上市申请。公司2023年至2025年收入从12.54亿元增至59.37亿元,2026年前4个月实现扭亏为盈。(每日经济新闻)

6 钛和检测港股IPO招股书失效

7月30日,钛和检测于2026年1月30日递交的港股招股书满6个月失效。公司是中国独立第三方TIC服务排名第五的企业。(证券之星)

7 沃客非凡港股IPO申请失效

7月20日,跨境新零售企业沃客非凡于2026年1月20日递交的港股上市申请因届满6个月失效。公司在印尼3C配件市场占有率排名第一。发布网站:新时空

8 飞速创新发盈喜,上半年净利润预增超六成

7月21日,飞速创新(03355.HK)公告预计2026年上半年净利润约4.374亿-4.647亿元,同比增加约60.1%至70.1%,主要受益于全球AI算力集群建设需求释放。

9 红星冷链获机构连续增持

7月30日,红星冷链(01641.HK)获Income SPC在场内以每股均价4.55港元增持3.75万股。此前7月16日,该机构曾增持159.15万股。

# 宏观政策 #

1 九部门发文加强科技金融领域数据开发利用

7月,中国人民银行等九部门发文加强科技金融领域数据开发利用,拥有央行企业征信备案资质的企查查等市场化数据机构迎来场景扩容机遇。(证券日报)

2 青岛西海岸新区“持牌”供应链金融科技公司揭牌

7月30日,西海岸新区首家持牌供应链金融科技公司揭牌运营,以供应链金融科技和产融数字化为核心,通过应收账款融资、订单融资等产品为上下游企业提供服务。(腾讯新闻)

# 本周金句 #

1 “中国HTTP请求中机器流量已占59.2%,超过人类流量。”

——中国工程院院士、紫金山实验室首席科学家刘韵洁,7月29日《中国智能互联网发展报告(2026)》发布会


2 “我们把数据库完整地搬到了GPU上运行,性能比CPU快20到100倍,这在全球还没有第二家企业做到。”

——星环科技创始人、CEO孙元浩


3 “解决了‘单个智能体能不能跑’的问题后,现在正是布局产业智能体互联网的关键窗口期。”

——万联易达集团副总裁杜新凯

# 本周叙评 #

本周产业互联网呈现三大显著形势:其一,工业AI从“单点试验”走向“系统原生”。其二,IPO市场冰火两重天,产业互联网迎来价值重估。其三,数据要素成为AI时代的核心基础设施。企查查MCP服务同时被腾讯、阿里两大生态接入,标志着商业数据正在成为AI智能体的“标配”能力。AI+数据要素的新时代,产业互联网的竞争已从规模比拼转向智能体能力与生态协同能力的较量。谁能率先跑通“模型+数据+场景”的商业闭环,谁就能在下一轮洗牌中占据先机。

文章来源:亿邦动力

广告
微信
朋友圈

FAQ回顾

当前产业互联网行业有哪些发展新趋势?

当前产业互联网呈现三大核心趋势:一是工业AI从“单点试验”走向“系统原生”;二是IPO市场冰火两重天,行业迎来价值重估;三是数据要素成为AI时代核心基础设施,行业竞争已从规模比拼转向智能体能力与生态协同能力较量。

中国HTTP请求中机器流量占比是多少?

据中国工程院院士、紫金山实验室首席科学家刘韵洁公开数据,中国HTTP请求中机器流量已占59.2%,超过人类流量。当前大模型发展重心已从“拼参数”转向“重应用”,正深入渗透农业、制造、医疗等实体经济领域。

产业智能体互联网是什么?

产业智能体互联网是万联易达在2026世界人工智能大会上正式提出的概念,目标是构建可“自动驾驶”的产业运作体系,体系内每个企业将拥有代表自身利益、能自主运转的“企业经纪人智能体”。

工业AI领域有哪些新落地的产品?

近期工业AI领域多款产品落地:京东工业发布覆盖采、存、标、训、评、仿、测全环节的“具身智链”布局;卡奥斯推出面向工业开发者的低门槛AI智能开发平台COSMO Code,支持自然语言驱动编程,可一句话秒出应用。

这么好看,分享一下?

朋友圈 分享

APP内打开

+1
+1
微信好友 朋友圈 新浪微博 QQ空间
关闭
收藏成功
发送
/140 0