广告
加载中

单季度交付百万小时数据 银河通用投了这家数据公司

亿邦动力 2026-11-02 13:33
亿邦动力 2026/11/02 13:33

邦小白快读

EN
全文速览

总:先抓住融资、公司定位和数据交付这条主线。

1.简智机器人完成新一轮1亿美元融资,产业方包括上汽集团、马来西亚Eastworlds、百度集团、首程控股、GALBOT(银河通用),财务方包括光合创投、SUNHELM、中信金石投资、中证投、建银国际、华泰紫金。

2.公司成立于2025年5月,定位为面向具身智能、世界模型的数据与评测基础设施公司,构建大规模人类技能数据收集、Data Foundation Model训练与数据处理、真机集群评测验证的闭环。

3.产品Gen DAS Dex采集终端采用仿生轻量化外骨骼,支持人手23个自由度检测,指尖磁触觉传感灵敏度达0.05N,空间分辨率1mm,可捕捉抚摸、抓取、按压等物理反馈。

总:再看增长数据和发布计划。

1.公司数据产品已在全球多家顶尖玩家物理AI模型上完成大规模验证,有效数据量从单季度十万小时跃升至单季度百万小时。

2.按现有产能与交付规划,公司预计2027年实现单季度交付千万小时。

3.公司将在CoRL期间发布DAS 2.0旗下Ego 2.0与Dex 2.0,DEX 2.0按模型数据需求反向定义采集系统,支持无本体采集与机器人遥操作双模式。

总:品牌商可关注产业资本背书、产品研发节奏和验证数据带来的品牌可信度。

1.融资由全球产业资本与顶级财投共同参与,产业方包括上汽集团、马来西亚Eastworlds、百度集团、首程控股、GALBOT(银河通用),财务方包括光合创投、SUNHELM、中信金石投资、中证投、建银国际、华泰紫金,形成较强的品牌背书。

2.产品研发上,Gen DAS Dex采集终端采用仿生轻量化外骨骼,支持人手23个自由度检测,指尖磁触觉传感灵敏度达0.05N、空间分辨率1mm。

总:发布节点和行业趋势可作为品牌传播与市场判断素材。

1.公司将在CoRL期间发布DAS 2.0旗下Ego 2.0与Dex 2.0,DEX 2.0基于模型数据需求反向定义采集系统,涉及侧置仿生外骨骼、关节内磁编码器、半柔性全手电子皮肤、Ego第一视角、腕部视觉、头手硬同步、无本体采集与遥操作双模式。

2.行业此前处于萌芽期,今年下半年进入成长期,头部基座模型企业开始系统性布局通用物理AI,未来应用期场景应用方将入局。

3.数据产品在全球多家顶尖玩家物理AI模型上完成大规模验证,交付量从单季度十万小时升至百万小时,并预计2027年达单季度千万小时。

总:卖家应重点看增长市场、需求变化和机会提示。

1.物理AI行业从萌芽期迈入成长期,头部基座模型企业开始系统性布局通用物理AI,未来行业步入应用期后,场景应用方也将正式入局,数据需求随之改变。

2.需求方从Frontier Labs走向Industry,需求性质从训练走向生产,这意味着数据生产、评测和场景落地环节存在市场机会。

3.简智构建大规模人类技能数据收集、DFM训练与数据处理、真机集群评测验证闭环,并以发布即量产、持续迭代DFM、真机评测集群形成规模化验证能力。

总:交付放量与合作风险值得关注。

1.公司数据产品已在全球多家顶尖玩家物理AI模型上完成大规模验证,有效数据量从单季度十万小时跃升至百万小时,预计2027年单季度交付千万小时,显示需求与产能扩张信号。

2.融资由产业资本与财务资本共同参与,产业方包括上汽集团、马来西亚Eastworlds、百度集团、首程控股、GALBOT(银河通用),可作为合作方式参考。

3.风险提示在于行业正处于成长期起点,数据需求方和需求性质仍会变化,卖家需关注从训练走向生产后的交付要求与验证门槛。

总:工厂可关注采集终端的产品设计需求和生产制造启示。

1.Gen DAS Dex采集终端采用仿生轻量化外骨骼,支持人手23个自由度检测,接近人手生理全部自由度;指尖搭载高精度磁触觉传感,灵敏度达0.05N,空间分辨率1mm,可捕捉抚摸、抓取、按压等动作物理反馈。

2.DAS 2.0旗下Dex 2.0基于模型数据需求反向定义采集系统,而非单一传感器功能堆叠,涉及侧置仿生外骨骼、关节内磁编码器、半柔性全手电子皮肤、Ego第一视角、腕部视觉、头手硬同步,并支持无本体采集与机器人遥操作双模式。

总:商业机会和数字化推进可从数据基础设施闭环中理解。

1.简智定位为具身智能、世界模型的数据与评测基础设施公司,构建大规模人类技能数据收集、Data Foundation Model训练与数据处理、真机集群评测验证闭环。

2.行业进入成长期,头部基座模型企业系统性布局通用物理AI,未来场景应用方入局,数据需求从训练走向生产,带来数据采集、评测验证和规模化交付需求。

3.公司强调发布即量产、持续迭代DFM、真机评测集群规模化验证,数据交付从单季度十万小时升至百万小时,并预计2027年单季度千万小时,对制造端推进量产和数字化交付有启示。

总:服务商可关注行业趋势、新技术和客户痛点解决方案。

1.行业趋势上,物理AI此前处于萌芽期,入局者以创业公司新玩家为主;今年下半年进入成长期,头部基座模型企业开始系统性布局通用物理AI;未来应用期场景应用方将入局。

2.需求变化上,数据需求方从Frontier Labs走向Industry,需求性质从训练走向生产,客户对可落地数据生产和评测验证的需求上升。

总:新技术与解决方案集中在数据闭环和采集评测能力。

1.简智构建大规模人类技能数据收集、Data Foundation Model训练与数据处理、真机集群评测验证闭环,为下一代具身、世界模型训练、评测与场景应用提供底层支撑。

2.新产品包括Gen DAS Dex采集终端和DAS 2.0旗下Ego 2.0与Dex 2.0,涉及仿生轻量化外骨骼、23个自由度检测、磁触觉传感0.05N、1mm空间分辨率、侧置仿生外骨骼、关节内磁编码器、半柔性全手电子皮肤、第一视角、腕部视觉、头手硬同步、无本体采集与遥操作双模式。

3.数据产品已在全球多家顶尖玩家物理AI模型上完成大规模验证,交付量从单季度十万小时升至百万小时,预计2027年单季度千万小时,可作为服务方案验证与产能参考。

总:平台商可关注生态需求变化和合作切入机会。

1.行业从萌芽期进入成长期,头部基座模型企业开始系统性布局通用物理AI,未来应用期场景应用方将正式入局,平台可关注由此带来的数据、评测和场景应用需求。

2.数据需求方从Frontier Labs走向Industry,需求性质从训练走向生产,说明需求将从前沿研发扩展到产业生产,平台合作对象和需求类型可能随之改变。

总:能力验证、产能和风向规避是平台评估重点。

1.简智具备大规模人类技能数据收集、DFM训练与数据处理、真机集群评测验证闭环,并强调发布即量产、持续迭代DFM、真机评测集群规模化验证能力,可作为合作方能力评估参考。

2.公司将在CoRL期间发布DAS 2.0旗下Ego 2.0与Dex 2.0,DEX 2.0按模型数据需求反向定义采集系统,支持无本体采集与机器人遥操作双模式,体现采集系统与模型需求协同。

3.数据产品已在全球多家顶尖玩家物理AI模型上完成大规模验证,有效数据量从单季度十万小时升至百万小时,预计2027年单季度千万小时。平台需关注行业仍处成长期起点、数据需求变化以及产能兑现节奏。

总:研究者可关注产业阶段、需求结构变化和商业模式。

1.产业新动向是物理AI从萌芽期进入成长期,头部基座模型企业开始系统性布局通用物理AI,未来应用期场景应用方将入局。

2.新问题在于数据需求方从Frontier Labs走向Industry,需求性质从训练走向生产,数据需求随行业阶段变化而改变。

总:简智的闭环体系、代表主体和交付数据可作为研究样本。

1.商业模式上,简智定位为具身智能、世界模型的数据与评测基础设施公司,构建大规模人类技能数据收集、Data Foundation Model训练与数据处理、真机集群评测验证闭环,为下一代具身、世界模型训练、评测与场景应用提供底层支撑。

2.代表主体上,公司成立于2025年5月,创始人、CEO陈建兴曾任职于小鹏、Momenta等,是前Momenta算法高级总监,曾牵头主导Momenta数据飞轮量产技术路线规划;核心团队来自大疆、华为、Momenta、Deepmind、斯坦福等。投资方包括上汽集团、马来西亚Eastworlds、百度集团、首程控股、GALBOT(银河通用)等产业方,以及光合创投、SUNHELM、中信金石投资、中证投、建银国际、华泰紫金等财务方。

3.数据上,公司数据产品已在全球多家顶尖玩家物理AI模型上完成大规模验证,有效数据量从单季度十万小时升至百万小时,预计2027年单季度千万小时。产品与技术包括Gen DAS Dex采集终端、DFM、真机集群、DAS 2.0旗下Ego 2.0与Dex 2.0。

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

Summary: Focus first on the funding, company positioning, and data delivery storyline.

1. Jianzhi Robotics has closed a new $100 million funding round. Strategic investors include SAIC Motor, Malaysia's Eastworlds, Baidu, Shoucheng Holdings, and GALBOT (Galbot); financial investors include Guanghe Ventures, SUNHELM, CITIC Jinshi Investment, Zhongzheng Investment, CCB International, and Huatai Zijin.

2. Founded in May 2025, the company positions itself as a data and evaluation infrastructure provider for embodied AI and world models. It is building a closed loop covering large-scale human skill data collection, Data Foundation Model (DFM) training and data processing, and real-robot cluster evaluation and validation.

3. Its Gen DAS Dex data-collection terminal uses a lightweight biomimetic exoskeleton and detects 23 degrees of freedom of the human hand. Its fingertip magnetic tactile sensing achieves 0.05 N sensitivity and 1 mm spatial resolution, capturing physical feedback such as stroking, grasping, and pressing.

Summary: The growth data and launch plans are also key.

1. The company's data products have undergone large-scale validation on physical AI models from multiple leading global players, with effective data volume rising from 100,000 hours per quarter to 1 million hours per quarter.

2. Based on current capacity and delivery plans, the company expects to deliver 10 million hours per quarter in 2027.

3. During CoRL, the company will release Ego 2.0 and Dex 2.0 under DAS 2.0. DEX 2.0 defines its collection system backward from model data requirements and supports dual modes: non-embodied data collection and robot teleoperation.

Summary: Brands can look at industrial-capital backing, product R&D cadence, and the credibility created by validated data.

1. The round brings together global industrial capital and top-tier financial investors. Strategic investors include SAIC Motor, Malaysia's Eastworlds, Baidu, Shoucheng Holdings, and GALBOT (Galbot); financial investors include Guanghe Ventures, SUNHELM, CITIC Jinshi Investment, Zhongzheng Investment, CCB International, and Huatai Zijin, providing strong brand endorsement.

2. On product R&D, the Gen DAS Dex data-collection terminal uses a lightweight biomimetic exoskeleton and detects 23 degrees of freedom of the human hand. Its fingertip magnetic tactile sensing achieves 0.05 N sensitivity and 1 mm spatial resolution.

Summary: Launch timing and industry trends can serve as material for brand communications and market assessment.

1. During CoRL, the company will release Ego 2.0 and Dex 2.0 under DAS 2.0. DEX 2.0 defines its collection system backward from model data requirements and involves a side-mounted biomimetic exoskeleton, in-joint magnetic encoders, semi-flexible full-hand electronic skin, Ego first-person view, wrist vision, hardware-level head-hand synchronization, and dual modes of non-embodied data collection and teleoperation.

2. The industry was previously in an emerging phase and entered a growth phase in the second half of this year. Leading foundation-model companies have begun systematically building out general-purpose physical AI, and in the future application phase, use-case application providers will enter the market.

3. Its data products have undergone large-scale validation on physical AI models from multiple leading global players. Deliveries have risen from 100,000 hours per quarter to 1 million hours per quarter and are expected to reach 10 million hours per quarter in 2027.

Summary: Sellers should focus on the growth market, shifts in demand, and opportunity signals.

1. The physical AI industry is moving from an emerging phase into a growth phase. Leading foundation-model companies have begun systematically building out general-purpose physical AI, and once the industry enters its application phase, use-case application providers will formally enter the market, changing data demand.

2. Demand is shifting from Frontier Labs to industry, and the nature of demand is shifting from training to production. This creates market opportunities in data production, evaluation, and use-case deployment.

3. Jianzhi is building a closed loop covering large-scale human skill data collection, DFM training and data processing, and real-robot cluster evaluation and validation. It also builds scalable validation capability through products that are production-ready at launch, continuous DFM iteration, and real-robot evaluation clusters.

Summary: Delivery scale-up and partnership risks warrant attention.

1. The company's data products have undergone large-scale validation on physical AI models from multiple leading global players. Effective data volume has risen from 100,000 hours per quarter to 1 million hours per quarter and is expected to reach 10 million hours per quarter in 2027, signaling demand and capacity expansion.

2. The funding round includes both industrial and financial capital. Strategic investors include SAIC Motor, Malaysia's Eastworlds, Baidu, Shoucheng Holdings, and GALBOT (Galbot), which can serve as a reference for partnership approaches.

3. The risk is that the industry is only at the start of its growth phase, and both data demand sources and the nature of demand will continue to change. Sellers need to watch delivery requirements and validation thresholds as demand shifts from training to production.

Summary: Factories can focus on product design requirements for data-collection terminals and the manufacturing implications.

1. The Gen DAS Dex data-collection terminal uses a lightweight biomimetic exoskeleton and detects 23 degrees of freedom of the human hand, close to the full physiological range of the human hand. Its fingertips carry high-precision magnetic tactile sensors with 0.05 N sensitivity and 1 mm spatial resolution, capturing physical feedback from actions such as stroking, grasping, and pressing.

2. Dex 2.0 under DAS 2.0 defines its collection system backward from model data requirements rather than stacking individual sensor functions. It involves a side-mounted biomimetic exoskeleton, in-joint magnetic encoders, semi-flexible full-hand electronic skin, Ego first-person view, wrist vision, hardware-level head-hand synchronization, and supports dual modes of non-embodied data collection and robot teleoperation.

Summary: Commercial opportunities and digitalization progress can be understood through the data-infrastructure closed loop.

1. Jianzhi positions itself as a data and evaluation infrastructure provider for embodied AI and world models. It is building a closed loop covering large-scale human skill data collection, Data Foundation Model training and data processing, and real-robot cluster evaluation and validation.

2. The industry has entered a growth phase. Leading foundation-model companies are systematically building out general-purpose physical AI, and use-case application providers will enter in the future. Data demand is shifting from training to production, creating demand for data collection, evaluation and validation, and scaled delivery.

3. The company emphasizes products that are production-ready at launch, continuous DFM iteration, and large-scale validation through real-robot evaluation clusters. Data delivery has risen from 100,000 hours per quarter to 1 million hours per quarter and is expected to reach 10 million hours per quarter in 2027, offering lessons for manufacturers in advancing volume production and digitalized delivery.

Summary: Service providers can focus on industry trends, new technologies, and solutions to customer pain points.

1. On industry trends, physical AI was previously in an emerging phase, with entrants mainly new players from startups. In the second half of this year, it entered a growth phase, as leading foundation-model companies began systematically building out general-purpose physical AI. In the future application phase, use-case application providers will enter.

2. On demand shifts, data demand is moving from Frontier Labs to industry, and the nature of demand from training to production. Customers increasingly need deployable data production and evaluation and validation.

Summary: New technologies and solutions center on data closed loops and collection and evaluation capabilities.

1. Jianzhi is building a closed loop covering large-scale human skill data collection, Data Foundation Model training and data processing, and real-robot cluster evaluation and validation, providing underlying support for next-generation embodied-AI and world-model training, evaluation, and use-case applications.

2. New products include the Gen DAS Dex data-collection terminal and Ego 2.0 and Dex 2.0 under DAS 2.0. They involve a lightweight biomimetic exoskeleton, 23-degree-of-freedom detection, magnetic tactile sensing at 0.05 N, 1 mm spatial resolution, a side-mounted biomimetic exoskeleton, in-joint magnetic encoders, semi-flexible full-hand electronic skin, first-person view, wrist vision, hardware-level head-hand synchronization, and dual modes of non-embodied data collection and teleoperation.

3. Its data products have undergone large-scale validation on physical AI models from multiple leading global players. Deliveries have risen from 100,000 hours per quarter to 1 million hours per quarter and are expected to reach 10 million hours per quarter in 2027, which can serve as a reference for validating service solutions and capacity.

Summary: Platform providers can focus on shifts in ecosystem demand and entry points for cooperation.

1. The industry is moving from an emerging phase to a growth phase. Leading foundation-model companies have begun systematically building out general-purpose physical AI, and in the future application phase, use-case application providers will formally enter. Platforms can watch the resulting demand for data, evaluation, and use-case applications.

2. Data demand is shifting from Frontier Labs to industry, and the nature of demand from training to production. This indicates that demand will expand from frontier R&D to industrial production, and platform partners and demand types may change accordingly.

Summary: Capability validation, capacity, and risk avoidance are key platform evaluation priorities.

1. Jianzhi has a closed loop covering large-scale human skill data collection, DFM training and data processing, and real-robot cluster evaluation and validation. It emphasizes products that are production-ready at launch, continuous DFM iteration, and large-scale validation through real-robot evaluation clusters, which can serve as a reference for evaluating partners' capabilities.

2. During CoRL, the company will release Ego 2.0 and Dex 2.0 under DAS 2.0. DEX 2.0 defines its collection system backward from model data requirements and supports dual modes of non-embodied data collection and robot teleoperation, reflecting coordination between collection systems and model requirements.

3. Its data products have undergone large-scale validation on physical AI models from multiple leading global players. Effective data volume has risen from 100,000 hours per quarter to 1 million hours per quarter and is expected to reach 10 million hours per quarter in 2027. Platforms need to watch that the industry is still at the start of its growth phase, along with changes in data demand and the pace of capacity delivery.

Summary: Researchers can focus on industry stage, shifts in demand structure, and business models.

1. A new industry development is that physical AI is moving from an emerging phase into a growth phase. Leading foundation-model companies have begun systematically building out general-purpose physical AI, and use-case application providers will enter in the future application phase.

2. A new question is that data demand is shifting from Frontier Labs to industry, and the nature of demand from training to production. Data demand is changing with the industry's stage.

Summary: Jianzhi's closed-loop system, key actors, and delivery data can serve as research samples.

1. On the business model, Jianzhi positions itself as a data and evaluation infrastructure provider for embodied AI and world models. It is building a closed loop covering large-scale human skill data collection, Data Foundation Model training and data processing, and real-robot cluster evaluation and validation, providing underlying support for next-generation embodied-AI and world-model training, evaluation, and use-case applications.

2. On key actors, the company was founded in May 2025. Founder and CEO Chen Jianxing previously worked at XPeng and Momenta and was a former Senior Algorithm Director at Momenta, where he led planning for Momenta's data flywheel mass-production technology roadmap. The core team comes from DJI, Huawei, Momenta, DeepMind, Stanford, and other organizations. Investors include strategic investors such as SAIC Motor, Malaysia's Eastworlds, Baidu, Shoucheng Holdings, and GALBOT (Galbot), as well as financial investors such as Guanghe Ventures, SUNHELM, CITIC Jinshi Investment, Zhongzheng Investment, CCB International, and Huatai Zijin.

3. On data, the company's data products have undergone large-scale validation on physical AI models from multiple leading global players. Effective data volume has risen from 100,000 hours per quarter to 1 million hours per quarter and is expected to reach 10 million hours per quarter in 2027. Its products and technologies include the Gen DAS Dex data-collection terminal, DFM, real-robot clusters, and Ego 2.0 and Dex 2.0 under DAS 2.0.

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.

【亿邦原创】日前,简智机器人完成新一轮1亿美元融资,由全球产业资本与顶级财投共同参与,产业方包括上汽集团、马来西亚Eastworlds、百度集团、首程控股、GALBOT(银河通用),财务方包括光合创投、SUNHELM、中信金石投资、中证投、建银国际、华泰紫金。

简智机器人(GenRobot)成立于2025年5月,是一家面向具身智能、世界模型的数据与评测基础设施公司,已构建起大规模人类技能数据收集、Data Foundation Model训练与数据处理、真机集群评测验证的闭环,打造规模化、高质量、可落地的物理AI数据生产与评测体系,为下一代具身、世界模型的训练、评测与场景应用提供底层支撑。

创始人、CEO陈建兴,曾任职于小鹏、Momenta等公司,是前Momenta算法高级总监,曾牵头主导Momenta数据飞轮量产技术路线规划。核心团队来自大疆、华为、Momenta、Deepmind、斯坦福等大厂和科研机构。

在产品层面,该公司推出了Gen DAS Dex采集终端。该产品采用仿生轻量化外骨骼,支持人手23个自由度检测,接近人手生理全部自由度。指尖搭载高精度磁触觉传感,灵敏度达0.05N,空间分辨率为1mm,可捕捉抚摸、抓取、按压等动作带来的物理反馈。

此前,物理AI行业处于萌芽期,入局者以创业公司新玩家为主。然而自今年下半年起,行业已迈入成长期,头部基座模型企业开始系统性布局通用物理AI。未来随着行业步入应用期,场景应用方也将正式入局,数据需求也将随之发生改变:需求方从Frontier Labs走向Industry,需求性质从训练走向生产。

当下,行业正处于成长期的起点。简智基于“发布即量产”的硬核底层能力、持续迭代的数据基础模型(DFM)、真机评测集群构建的规模化验证能力。

简智还宣布,将于CoRL(机器人学习会议)期间正式发布DAS 2.0旗下Ego 2.0与Dex 2.0两款新品。DEX 2.0基于模型数据需求反向定义采集系统,而非单一传感器的功能堆叠;从侧置仿生外骨骼、关节内磁编码器到半柔性全手电子皮肤,从Ego第一视角、腕部视觉到头手硬同步,同步支持无本体采集与机器人遥操作双模式。

简智方面表示,公司的数据产品已在全球多家顶尖玩家的物理AI模型上完成大规模验证,交付的有效数据量已经从单季度十万小时跃升至单季度百万小时,按照现有数据产能与交付规划,公司预计在2027年将实现单季度交付千万小时。

本文首发于 亿邦动力 官方网站

文章来源:亿邦动力

广告
微信
朋友圈

FAQ回顾

简智机器人是一家什么公司?

简智机器人(GenRobot)成立于2025年5月,是一家面向具身智能和世界模型的数据与评测基础设施公司。公司已构建大规模人类技能数据收集、Data Foundation Model训练与数据处理、真机集群评测验证的闭环,为下一代具身模型、世界模型的训练、评测与场景应用提供底层支撑。

简智机器人的融资规模和投资方有哪些?

简智机器人完成新一轮1亿美元融资。产业投资方包括上汽集团、马来西亚Eastworlds、百度集团、首程控股和GALBOT(银河通用);财务投资方包括光合创投、SUNHELM、中信金石投资、中证投、建银国际和华泰紫金。

简智机器人的Gen DAS Dex采集终端有哪些技术特点?

Gen DAS Dex采集终端采用仿生轻量化外骨骼,支持人手23个自由度检测,接近人手生理全部自由度。其指尖搭载高精度磁触觉传感,灵敏度达0.05N,空间分辨率为1mm,可捕捉抚摸、抓取、按压等动作带来的物理反馈。

物理AI数据行业的需求和产能正在发生什么变化?

物理AI行业已从萌芽期迈入成长期,头部基座模型企业开始系统性布局通用物理AI。未来应用期,需求方将从Frontier Labs走向Industry,需求性质从训练走向生产。简智机器人的有效数据交付量已从单季度十万小时升至单季度百万小时,并预计2027年实现单季度交付千万小时。

简智机器人DAS 2.0将推出哪些新品,Dex 2.0有何设计特点?

简智机器人将在CoRL(机器人学习会议)期间发布DAS 2.0旗下的Ego 2.0与Dex 2.0。Dex 2.0基于模型数据需求反向定义采集系统,而不是单一传感器功能堆叠;整合侧置仿生外骨骼、关节内磁编码器、半柔性全手电子皮肤、Ego第一视角、腕部视觉和头手硬同步,并支持无本体采集与机器人遥操作双模式。

这么好看,分享一下?

朋友圈 分享

APP内打开

赞 +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