Zhiwang Future Demonstrates Armstrong Pro Robot at CeMAT with Focus on Real-World Warehouse Economics
具身智能的“算账时代”:这家企业在仓库里死磕“1公分”与“1比1”替人
At the CeMAT exhibition, Zhiwang Future showcased its second-generation Armstrong Pro wheeled dual-arm robot performing autonomous navigation and item picking in a dense pharmaceutical warehouse. CEO Sun Junkai highlighted that while early demonstrations attracted curiosity, current industry focus has shifted to practical deployment and economic viability. The company addresses physical constraints like a 1-centimeter reach limitation through hardware iteration and a hybrid AI architecture combining world action models, skill libraries, and reinforcement learning. Although initial failure rates range from 1% to 3%, the system allows retries before human intervention. Zhiwang Future aims for a 1:1 replacement ratio with workers over time, targeting hundreds of units deployment by next quarter in existing warehouses without structural modifications.
Source: Chinese robotics — Deployment discovery · Read original article ↗
English translation unavailable. Choose the original article to read it.
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Control
- Reported autonomousOpen source S4
Reported numbers
sku_count
100,000
- paper: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
受访者/供图) 中经记者 吴静 卢志坤 北京报道 11月3日,上海新国际博览中心,亚洲国际物流技术与运输系统展览会(CeMAT)如期开幕。
Open source S3
在具身智能企业智往未来的展台上,观众正通过iPad体验“下单”——随着指令下发,其二代机器人轮式双臂机器人Armstrong Pro自主导航至密集货架,精准抓取指定物品并放入料箱。
Open source S4
” 即便到现在,机器人的整体失败率依然在1%到3%。
Open source S14
面对失败,系统允许连续尝试两次,若无果则呼叫人工兜底,随后机器人转入下一个任务。
Open source S15
“失败一次就要重来一次,相当于浪费一次有效拣选时长,粗略估计每小时会影响数十件产能。
Open source S16
只有在系统积累了10万个SKU的经验后,当第10万零1个新SKU出现时,学习时间才能缩短到“小时/分钟级别”。
Open source S20
在上述药企的算账逻辑里,一个工人三年的综合成本在几十万元左右,而机器人的生命周期是五年。
Open source S24
Source:Chinese robotics — Deployment discovery · cj.sina.com.cn