Daxiao Robotics Deploys Dual-Brain Architecture for Autonomous Hotel Laundry Services
Overview
Daxiao Robotics has deployed its dual-brain architecture, combining Mundo world models, GPT-Astra general models, and RSI recursive self-improvement strategies, into a hotel laundry scenario. The system uses digital twin simulations for pre-training before executing tasks on real machines. Test data indicates that after optimizing action logic into skills, execution time dropped from over 30 minutes to approximately 18.648 seconds per cycle. Following 165 rounds of real-machine experiments and about 80 version iterations, the robot successfully improved performance in complex tasks like retrieving clothes from dryers.
Generated from attributed reports · Updated 21 hours ago
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
Control: reported autonomous
“东方网记者曹磊11月10日报道：从客房取衣、穿行走廊、操作洗衣机与烘干机，到烘干装篮再送回客房，一套完整酒店洗衣服务流程，正由机器人自主完成。”
Exact source · revision 2Source owner not reported
Reported quantity · Reported duration: 30 other · Basis not reported
“测试数据显示，未经优化时单次新操作耗时超 30 分钟，沉淀为固化技能后一轮执行仅需约 18.648 秒，提速约 97 倍。”
Exact source · revision 2Source owner not reported
Reported quantity · elapsed_seconds: 18.648 other · Basis not reported
“测试数据显示，未经优化时单次新操作耗时超 30 分钟，沉淀为固化技能后一轮执行仅需约 18.648 秒，提速约 97 倍。”
Exact source · revision 2Source owner not reported
Reported quantity · trials: 165 other · reported trials
“以烘干机取衣为例，经过 165 轮真机实验、约 80 个版本迭代，牛仔裤残留筒内比例显著下降，充分证明真实物理反馈是机器人能力升级的关键动力。”
Exact source · revision 2Source owner not reported
Report timeline
Follow attributed reports and material updates.
- Robotics — Chinese web discoverySignalDaxiao Robotics Deploys Dual-Brain Architecture for Autonomous Hotel Laundry Services
Daxiao Robotics has deployed its dual-brain architecture, combining Mundo world models, GPT-Astra general models, and RSI recursive self-improvement strategies, into a hotel laundry scenario. The system uses digital twin simulations for pre-training before executing tasks on real machines. Test data indicates that after optimizing action logic into skills, execution time dropped from over 30 minutes to approximately 18.648 seconds per cycle. Following 165 rounds of real-machine experiments and about 80 version iterations, the robot successfully improved performance in complex tasks like retrieving clothes from dryers.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).