News
News
Home / News
Humanoid robot in cinema autonomously sells over 1,000 cups of popcorn in 14 hours
source:羊城晚报-PEARL| 发表时间:2026-01-05 22:55

On January 4th, 2026, Dobot Robotics achieved the first fully autonomous operation of a humanoid robot in a complex commercial environment. Its fully autonomous humanoid robot has been deployed at the K11 Art House Cinema in Shenzhen, capable of working continuously for 14 hours a day, independently producing and selling over 1,000 cups of popcorn without any human help.

It is reported that the robot is equipped with Dobot's self-developed Dobot-VLA end-to-end model, which integrates vision, language, and action capabilities. The system can translate task instructions into continuous, generalizable actions, allowing the robot to autonomously perform the entire workflow, including recognition, grasping, filling, and delivery in the dynamic, open environment of a cinema counter. It also achieves millimeter‑level grasping control and flexible obstacle avoidance.

This application is supported by Dobot's self‑developed "big brain and small brain" coordination system, which provides autonomous task planning and real‑time error correction. When encountering unexpected disturbances such as spilled materials or moved containers, the robot can independently identify the situation, diagnose the problem, and dynamically re-plan its actions, maintaining stable operation with "zero errors."

Industry experts noted that this case marks a critical breakthrough for humanoid robots in non‑standardized scenarios, demonstrating deep coordination between the "brain" making decisions and the "hands" executing tasks. It advances embodied intelligence technology from the laboratory to real‑world commercial settings, offering an important model for differentiated application in the industry.

人形机器人“入职”影院,14小时可自主售卖超千杯爆米花

2026年1月4日,越疆机器人实现首例人形机器人在复杂商业环境中全自主运营的落地应用,其全自主人形机器人成功进驻深圳K11 Art House电影院,单日可连续工作14小时,独立完成超过1000杯爆米花的制作与售卖,全程无需人工干预。

据悉,该机器人搭载自研Dobot-VLA端到端模型,融合视觉、语言与动作能力,能将任务指令转化为连续可泛化的动作,在动态开放的影院柜台环境中自主完成识别、抓取、装填、递送等全流程作业,并实现毫米级抓取控制和灵活避障。

此次应用依托于越疆自研“大小脑”协同系统,具备自主任务规划与实时纠错能力。在遇到物料洒落、容器移动等突发干扰时,机器人可自主识别状态、诊断问题并动态重规划,保持“零失误”稳定运行。

业内相关人士指出,该案例标志着人形机器人在非标准化场景中实现关键突破,体现了“大脑”决策与“双手”执行的深度协同,推动具身智能技术从实验室走向真实商业场景,为行业差异化应用落地提供了重要示范。

文、图|记者 潘亮 实习生 黄婧彤
译 | 黄宇洁
审校 | 赵凡

  • APEC 'China Year' sees first tech-focused auto show this May
    As the year 2026 marks the beginning of China's 15th Five-Year Plan, the Guangdong-Hong Kong-Macao Greater Bay Area International Auto Show enters a new chapter, celebrating its 30th anniversary.
    2026-01-05 22:53:52
  • 42,000 Visits on New Year’s Eve: Dongmen Market in Foshan Draws Holiday Crowds
    ​On December 31, 2025, Dongmen Market in Foshan attracted 42,000 visitors, generating 563,000 yuan in consumption.
    2026-01-05 21:17:05
  • Guangdong’s Own “Route 318”: Discover the Beauty Along 218 Highway
    ​Stretching 218 kilometers in the shape of "8", the highway in Huizhou’s Nankun Mountain-Luofu Mountain Pioneer Area unfolds like a flowing landscape painting.
    2026-01-05 21:17:05
  • New Year in Foshan's Historicbuilding: Shadow Puppetry Meets Lingnan Architecture
    During the New Year Holiday,Lingnan Xintiandi in Foshan welcomed visitors to experience a garden tour and theater performances.
    2026-01-05 21:17:05
Scan to Share