Workflow
Robotics
icon
Search documents
具身智能领域的行业周期有多久?
具身智能之心· 2025-06-22 03:59
Core Viewpoint - The article discusses the development cycles of autonomous driving and embodied intelligence, suggesting that the latter may achieve commercialization faster due to anticipated breakthroughs in algorithms and data [1]. Group 1: Industry Development - The autonomous driving industry has been scaling and commercializing for nearly 10 years since 2015, while the robotics industry has been evolving for many years, with expectations for significant advancements in the next 5-8 years [1]. - Companies like Zhiyuan and Yushu are preparing for IPOs, which could greatly invigorate the entire industry [1]. Group 2: Community Building - The goal is to create a community of 10,000 members within three years, focusing on bridging academia and industry, and providing a platform for rapid problem-solving and industry influence [1]. - The community aims to facilitate technical exchanges and discussions on academic and engineering issues, with members from renowned universities and leading robotics companies [8]. Group 3: Educational Resources - A comprehensive entry route for beginners has been organized within the community, including various learning paths and resources for those new to the field [2]. - For those already engaged in research, valuable industry frameworks and project proposals are provided [4]. Group 4: Job Opportunities - The community continuously shares job postings and opportunities, contributing to the establishment of a complete ecosystem for embodied intelligence [6]. Group 5: Knowledge Sharing - The community has compiled a wealth of resources, including over 40 open-source projects, nearly 60 datasets related to embodied intelligence, and mainstream simulation platforms [11]. - Various learning routes are available, covering topics such as reinforcement learning, multi-modal models, and robotic navigation [11].
蚂蚁集团频频出手具身智能,新投资宇树科技、钛虎、灵心巧手
Nan Fang Du Shi Bao· 2025-06-21 08:22
蚂蚁集团近一周内新增三笔具身智能赛道投资,成为出手最活跃的互联网大厂之一。这三笔投资涵盖人 形机器人本体厂商宇树科技、钛虎机器人,以及灵巧手零部件公司灵心巧手。 另据记者了解,蚂蚁近期还投资了位于上海的钛虎机器人。6月18日,这笔投资完成工商信息变更,投 资主体为蚂蚁全资持有的上海云玡企业管理咨询有限公司。公开信息显示,钛虎机器人成立于2020年8 月,主要研发高性能一体化关节,并在此基础上推出了多款双足和轮式的人形机器人。这家公司此前已 获梅花创投、经纬创投等知名市场化机构押注。 6月16日,蚂蚁全资持有的上海云玚企业管理咨询有限公司,成为灵心巧手的新股东。这是灵心巧手的 天使轮融资,和蚂蚁一同加入的还有中金资本。此前4月,灵心巧手完成种子轮融资,红杉种子基金等 领投,金额超亿元。灵巧手是人形机器人上岗干活的关键零部件,技术研发难度极高。灵心巧手公司目 前主打的多款灵巧手款式,价格从0.88万至9.99万元不等。 此前,蚂蚁已在具身智能赛道布局了星海图、星尘智能等创业公司。2024年11月,星海图完成超2亿元 的Pre-A轮融资,蚂蚁即为领投方之一。今年2月,蚂蚁又独家领投了星海图3亿元的A轮融资。随后的4 ...
“具身智能”成国家未来重点培育产业 训练师岗位成“全人才梯度吸纳场”
Yang Shi Wang· 2025-06-21 08:09
央视网消息:今年3月,"具身智能"首次被写入两会政府工作报告,成为国家未来重点培育产业,其中,人形机器人是具身智 能的典型代表。近年来,人形机器人越来越灵活,应用场景也越来越广泛。产业升级发展的背后离不开一个新职业——具身智能 训练师。他们是如何让机器人变得越来越"能干"呢?一起去上海的一家人形机器人训练场看一看。 采集到的视频数据,还要转化为机器人能"看得懂"的精确数据,这就来到了机器人训练的第二步,数据审核。具身智能训练 师宋子文的工作是进行机器人数据审核,他每天要审核上百条数据,不到一年时间,他和小伙伴们就审核了400多万条的真机数 据。 身着动作捕捉设备,手把手教授机器人的这位"动作老师",是朱绵锐。 他正在进行的工作是采集数据,这也是机器人训练的 第一步。根据任务难易程度,一个采集师每天会产出50-200条数据。 宋子文口中提到的算法同事就是杨树开。现在他正在机器人上测试模型的表现,这个模型正是昨天他收到数据后,用算法训 练而来。 数据采集岗的几个小伙儿来自湖北,是计算机专业的应届毕业生。 具身智能训练师并非某一个人,想要完成机器人的训练,要3个主要岗位共同配合。 已经具备基础擦桌子能力的机器人,想 ...
星动纪元推出超拟人服务机器人星动Q5,已签订超百单意向订单
IPO早知道· 2025-06-21 01:52
唯一一家清华大学占股的人形机器人企业。 本文为IPO早知道原创 作者| Stone Jin 微信公众号|ipozaozhidao 据IPO早知道消息,清华系具身智能企业「星动纪元」日前推出了一款超拟人服务机器人星动Q5。 基于曼妙"小蛮腰"设计 ,被业内戏称为机器人界"小腰精" 。 不妨补充一点, 成立于2023年8月的星动纪元由清华大学交叉信息研究院孵化,是唯一一家清华大 学占股的人形机器人企业。本次星动Q5的发布,某种程度上标志着服务机器人行业正式迈入 "交互 体验升级" 新阶段,并已与海尔智家、世纪金源、天悦酒店等来自制造、连锁零售、连锁酒店、文 旅、养老医疗等业务的企业达成战略签约,意向订单超百单 。 带着服务机器人从拼"功能" 迈入卷"体验" 时代。 对比来看,传统服务机器人动作僵硬、人机交流冰冷,很难引起人们主动与其交流的意愿。星动Q5 靠全身44个高自由度,打造超拟人的外观和动作系统。独特设计"小蛮腰"不但让重心下移,移动更 稳,还让它的身形更贴近真人,高自由度使它能够实现扭腰、握手、摆臂、转头等接近人类肢体的动 作,让人机交互更加自然生动,尤其和老人、小孩互动时,亲和力直接拉满。 搭配拟人化语 ...
Robot Delivery Service Plans US and Europe Expansion
Bloomberg Technology· 2025-06-20 19:23
The money. Was it for Zach. Well, we really want to expand the service to more areas, so we're building the best autonomous vehicle for our urban environment.So we're operating in Los Angeles, Chicago, Miami, Helsinki, Finland. We have a lot more cities coming up soon. But, you know, we want to build the best vehicle to move goods around cities to make it much cheaper for our customers and for merchants.Okay. So it's all about U.S. markets. Is there an international focus as well here.Yes, In the US we've b ...
Amazon says new warehouse robot can 'feel' items, but won't replace workers
CNBC· 2025-06-20 16:00
We're here at this giant 2.6% million square foot fulfillment center in Spokane, Washington, where 3.5% million items are handled by 2500 employees. Now, Amazon says their jobs are about to get easier and safer with a new robot called Vulcan. There's three of them, these big red systems next to me that can move items up and down and stuff things into just the right spot in tall yellow bins before they move on to be boxed up.Of the 1 million items here, Amazon says Vulcan can handle 75% of them. It does well ...
4 Top Robotics Stocks to Watch in the Second Half of 2025
ZACKS· 2025-06-20 15:40
Industry Overview - The robotics industry is at a significant inflection point, driven by AI integration, with a projected annual growth rate of 15.1%, leading to a global market valuation of $169.8 billion by 2032 [2] - The global industrial robotics market is expected to grow from $87.1 billion in 2024 to $162.7 billion by 2030, reflecting an 11% CAGR [3] - The medical service robots sector is anticipated to reach $51.9 million by 2030, showcasing advancements in surgical robotics [4] - Collaborative robots (cobots) are projected to see a 6,100% increase in sales from 2025 to 2045, particularly in food and beverage packaging [5] - The defense and space exploration sectors are leveraging robotics for autonomous systems, with companies like Palantir and SpaceX leading the way [6] Company Highlights - Advanced Micro Devices (AMD) is enhancing its robotics portfolio with adaptive computing solutions, including the Kria SOM platform for AI sensor data processing and real-time digital twin simulations [9][10] - Tesla is expanding its robotics initiatives beyond automotive with the Optimus humanoid robot, targeting mass production of 50,000-100,000 units by 2026, and plans to send Optimus to Mars by 2026 [11][12] - Cadence Design Systems is a key player in robotics, providing AI processors and simulation tools that optimize robotic system design and performance [13] - Serve Robotics is innovating in the delivery sector with AI-powered sidewalk robots, completing tens of thousands of deliveries and planning to deploy up to 2,000 robots across U.S. markets [14][15]
晚点独家丨宇树 C 轮落定,中移动、腾讯、锦秋、阿里、蚂蚁、吉利共同领投,估值超 100 亿
晚点LatePost· 2025-06-20 13:20
以下文章来源于晚点科技 ,作者晚点团队 据了解,宇树科技已于近期完成了去年底开启的 C 轮融资,由中国移动旗下基金、腾讯、锦 秋、阿里、蚂蚁和吉利资本共同领投,宇树的大部分老股东跟投。宇树投前估值目前超过 100 亿元人民币。 在去年底至今的 C 轮融资期间,宇树经历了春节晚会出圈爆火、公司股改等系列事件。据《晚 点 LatePost》此前对这轮融资的跟踪了解:今年 1 月初,中国移动、蚂蚁和一些大型产业方就 在积极洽谈,几家主要潜在投资方计划出资 1 亿-3 亿元不等。而进入今年 2 月后,宇树 C 轮 投后估值逐渐上涨,融资结构也相应变化。 据多方信息,有原本希望参与 C 轮融资的潜在产业投资方后来也购买了宇树的部分老股。 晚点科技 . 晚一点,好一点 宇树的产品正成为具身研发基础设施。 文 丨 贺乾明 李梓楠 编辑 丨 程曼祺 宇树的估值目前并不是中国具身市场里最高的。"宇树的(正式)估值在十名开外。" 一位宇树的早期 投资人不久前曾对我们感叹,当时宇树的 C 轮仍未最终落定。 论知名度、影响力和实际出货量,宇树的机器狗和人形机器人都位居中国乃至世界前列。据多方行业 信息,今年上半年,宇树部分老股交易的 ...
人形机器人“闹展会”,量产易、应用难
3 6 Ke· 2025-06-20 12:15
当AI大模型以星火燎原之势渗透至千行百业,作为其重要落地载体的具身智能,正以"现实版钢铁侠"的姿态,成为科技展会中"最靓的仔"。 从通信技术中来,往通信世界里去 人形机器人向来是科技展会中最吸睛的存在。 一大早,智元机器人展台早已挤满前来参观的观众。远征A2手持毛笔,一笔一画写着"福"字;灵犀X2不仅用"内心戏"模式与观众互动,还向观众表演了 一段太极拳。这些能力的背后,既有智元对模型架构的创新构建,也少不了通信技术的支持。 智元打造了"本体—小脑—大脑"的软硬件技术架构,让人形机器人实现了运动智能、交互智能和作业智能。"我们将一些基本能力,比如手脚运动,做在 本体和小脑中,使机器人在断网的情况下,也能实现基本操作。"智元机器人首席运营官邱恒告诉《IT时报》记者,"大脑"作为人形机器人智慧的关键, 由云平台+具身算法构建而成,通信技术被运用其中。"有了通信技术的加持,就像给人形机器人配备了一台可以实时获取信息的手机,联网后能获得更 多智慧,一些复杂问题也将交由云端处理,交互就会更加'聪明'。" 具备这些能力后,人形机器人将走进通信场景。智元旗下的远征A2、精灵G1、灵犀X2等多款机器人将进入展厅、营业厅、机房 ...
突破开放世界移动操作!首个室内移动抓取多模态智能体亮相,微调模型真实环境零样本动作准确率达 90%
机器之心· 2025-06-20 11:59
在家庭服务机器人领域,如何让机器人理解开放环境中的自然语言指令、动态规划行动路径并精准执行操作,一直是学界和工业界的核心挑战。 近日,上海人工智能实验室联合新加坡国立大学、香港大学等机构的研究团队,提出了 " OWMM-Agent " 具身智能体——首个专为开放世界移动操作 (OWMM)设计的多模态智能体 (VLM Agent) 架构,首次实现了全局场景理解、机器人状态跟踪和多模态动作生成的统一建模。 同时该工作通过仿真器合成智能体轨迹数据,微调了针对该任务的多模态大模型 OWMM-VLM,在真实环境测试下,该模型零样本单步动作预测准确率达 90%。 论文链接:https://arxiv.org/pdf/2506.04217 Github 主页:https://github.com/HHYHRHY/OWMM-Agent 一、问题背景介绍:开放语义下的移动抓取任务 传统移动抓取机器人在家庭场景处理 "清理餐桌并将水果放回碗中" 这类开放指令时,往往需要依赖预先构建的场景 3D 重建或者语义地图,不仅耗时且 难以应对动态环境。OWMM 任务的核心难点在于: 二、OWMM-Agent:用 VLM 重构机器人 "大脑 ...