Workflow
机器人工程化
icon
Search documents
机器人从苏州走到上海,106公里没摔跤
第一财经· 2025-11-21 10:41
" 机器人晚上 10 点才能开始走,早上 6 点前要收工。 " 从苏州金鸡湖到上海外滩,智元旗下的机器人远征 A2 完成了 106.286 公里的行走,也成为了吉尼斯 世界纪录官方认证的 " 人形机器人行走最远距离 " 世界纪录保持者。尽管没有摔一次跤,但机器人进入 真实世界,还是遇到了不小的挑战:走人行道还是走非机动车道?没有补给点设施怎么换电?要不要全 自主?万一碰到了人怎么办? 2025.11. 21 本文字数:2125,阅读时长大约4分钟 作者 | 第一财经 乔心怡 第一财经记者采访了支持本次远征 A2 百公里行走的幕后团队了解到,本次行走是机器人融入人类真实 生活的一个尝试。出于安全性考虑,机器人本次的行走只能在深夜进行。 王闯透露,半自主是团队考虑到实际情况之后的妥协策略。 " 虽然夜间人会少一点,但实际路况还是比 较复杂,而且汽车有规则,有些地方没有非机动车道,机器人没有规则,我们考虑到安全性,就没有选 择全自主方案。 " 实时变化着的真实世界考验机器人背后庞大工程体系的稳定性、鲁棒性与技术边界。 " 机器人已经走出 实验室,开始交付,但在真实世界的商用部署里会遇到什么问题,不去试就不会知道。 ...
机器人从苏州走到上海,106公里没摔跤,但卡在“规则”与“补给”
Di Yi Cai Jing· 2025-11-21 10:00
机器人走106公里,换电是个大问题。 根据远征A2的日常续航情况,团队的工作人员透露,A2的电池续航在站立状态下能持续3小时,"行走的话,能够走2 小时"。但在本次行走的实际过程中,为了保障安全,工作人员透露,会根据实际状况调整。"毕竟路上什么人都有, 走到最低电量才换电会有安全隐患。" 真实世界考验机器人背后庞大工程体系的稳定性、鲁棒性与技术边界。 "机器人晚上10点才能开始走,早上6点前要收工。" 从苏州金鸡湖到上海外滩,智元旗下的机器人远征A2完成了106.286公里的行走,也成为了吉尼斯世界纪录官方认证 的"人形机器人行走最远距离"世界纪录保持者。尽管没有摔一次跤,但机器人进入真实世界,还是遇到了不小的挑 战:走人行道还是走非机动车道?没有补给点设施怎么换电?要不要全自主?万一碰到了人怎么办? 第一财经记者采访了支持本次远征A2百公里行走的幕后团队了解到,本次行走是机器人融入人类真实生活的一个尝 试。出于安全性考虑,机器人本次的行走只能在深夜进行。 实时变化着的真实世界考验机器人背后庞大工程体系的稳定性、鲁棒性与技术边界。"机器人已经走出实验室,开始 交付,但在真实世界的商用部署里会遇到什么问题,不去 ...
“全员格斗”背后:别让流量焦虑掩盖机器人工程化难题
第一财经· 2025-05-26 14:51
Core Viewpoint - The robotics industry is evolving rapidly, with companies like Yushu and Zhongqing showcasing advanced capabilities through combat competitions, reflecting a collective anxiety within the sector as it seeks to capture attention and validate technology [1][2]. Group 1: Industry Trends - The robotics sector is moving from basic functionalities like walking to more complex demonstrations such as dancing and fighting, indicating a shift in focus towards more engaging displays [1][2]. - The emergence of combat competitions is not merely a technical showcase but a strategic move to capture public interest amid a backdrop of technological challenges and unclear business models [2][3]. Group 2: Technical Insights - Yushu's G1 robot demonstrated notable performance in a recent combat event, showcasing its hardware and algorithm stability despite some operational challenges [3][4]. - Combat scenarios serve as a comprehensive test for robots, evaluating their mechanical structure, motion control, sensor integration, and intelligent decision-making capabilities, which are crucial for real-world applications [4]. Group 3: Industry Concerns - There are concerns that the focus on attention-grabbing events like combat competitions may overshadow significant engineering challenges, such as hardware-software integration, power consumption, and long-term stability [4].
“全员格斗”背后:别让流量焦虑掩盖机器人工程化难题
Di Yi Cai Jing· 2025-05-26 03:39
Core Viewpoint - The robot fighting competitions are not merely technical showcases but represent a competition for attention within the industry [1][4][5]. Group 1: Industry Trends - The complexity of robotic demonstrations is increasing, moving from walking to running and now to fighting, as companies seek to capture attention [4][5]. - Companies like Yushun and Zhongqing are actively participating in robot fighting events, indicating a shift in focus towards more engaging applications [1][4]. - The industry is experiencing collective anxiety, as evidenced by the proliferation of fighting events, which serve as an extension of technology showcases [4][5]. Group 2: Technical Insights - Yushun's G1 robot demonstrated notable performance in a recent competition, showcasing its hardware and algorithm stability despite some operational challenges [5][6]. - The ability of G1 to maintain balance during high-intensity confrontations relies on AI reinforcement learning and joint perception [6]. - Fighting scenarios can comprehensively test a robot's mechanical structure, motion control, sensor fusion, and intelligent decision-making capabilities, which are crucial for real-world applications [6]. Group 3: Industry Concerns - There are concerns regarding whether these competitions are a natural evolution of technology or merely a marketing gimmick to attract attention [5][6]. - The focus on attention-grabbing events may overshadow significant engineering challenges, such as hardware-software integration, power consumption, and long-term stability [6].