返回新聞
產品展示AI Understanding 簡報

Newbility 推出用於自主物料搬運的混合人形機器人“Billy”

韓國實體人工智慧公司 Newbility 展示了其新型混合人形機器人 Billy,它可以自主導航並舉起重達 15 公斤的物體,旨在在食品、物流和製造領域進行商業部署。

4 min readRead the linked source
Source-provided image accompanying Newbility unveils hybrid humanoid robot ‘Billy’ for autonomous material handling
來源參考來源記錄
出版商
mk.co.kr
來源連結
mk.co.krhttps://www.mk.co.kr/en/business/12163655
來源類型
連結來源-主要來源狀態尚未確定。
背景60 秒內了解這一點

從這裡開始

關鍵術語

推理
經過訓練的模型產生預測或輸出的運行時階段。
精確度
實際正確的預測陽性的比例。
計算
訓練和運行模型所需的處理資源,通常以 FLOPS 或 GPU 小時來衡量。
測試一下自己AI 代理測驗

發生了什麼事

Newbility introduced a hybrid humanoid robot named Billy at a media preview in Seoul on September 29. The robot combines an upper torso with arms and hands and a lower chassis with four wheels, allowing it to drive autonomously and manipulate items. In the demonstration, Billy moved from an elevator to a shelf three to four metres away, picked up a box, and placed it in a rear‑mounted loading box. The robot can lift objects under 15 kg directly and carry up to 60 kg in its rear box. Newbility said it has accumulated more than 140,000 km of autonomous driving data from 400 robots deployed at over 150 sites, and plans to integrate this experience with manipulation data to train a unified model. The company announced upcoming proof‑of‑concept trials with large food manufacturers, manufacturers, logistics firms and transportation complexes, including a PoC with CJ CheilJedang slated for the second half of the year.

At the KT&G Sangsang Planet venue in Seongsu‑dong, Seoul, Newbility’s CEO and CTO Kang Ki‑hyuk highlighted Billy’s ability to understand space and directly manipulate parts needed for a task. The robot’s upper body can extend to a maximum height of 149.7 cm, and its rear loading box can hold up to 60 kg of cargo.

Newbility’s product owner Kim Rak‑hee estimated that Billy can operate for four to six hours per shift, effectively replacing a human worker for that duration, aside from rest periods. The company emphasized that unlike conventional collaborative robots that perform repetitive tasks at fixed locations, Billy can handle multiple tasks within a single workflow, offering greater versatility.

The firm reported that its fleet of 400 robots has generated roughly 145 million visual and spatial perception data points annually, amounting to over 140,000 km of autonomous driving experience. This data is intended to train a single model that merges navigation and manipulation capabilities, a concept the company refers to as “space intelligence.”

來源詳情: mk.co.kr ↗

為什麼這很重要

The launch marks a notable step toward physically embodied AI that can bridge the gap between fixed‑position collaborative robots and fully mobile automation. By combining autonomous navigation with manipulation, Billy could reduce labor costs and increase flexibility in environments where processes are not fully automated, such as food‑processing lines that still rely on human operators to move items between stations. Newbility’s claim of leveraging extensive real‑world driving data to improve spatial intelligence suggests a move toward more generalized robot cognition, which could accelerate adoption across diverse industries. If the upcoming PoCs succeed, the robot could be deployed at scale, potentially reshaping material‑handling workflows and prompting competitors to develop similar hybrid platforms.

Hybrid mobility and manipulation could address a persistent bottleneck in many factories where automation exists but human workers still bridge process gaps. By automating these bridging tasks, companies may improve throughput and reduce reliance on manual labor, which is increasingly scarce in some regions.

The integration of large‑scale perception data into a unified model may set a precedent for how physical AI systems learn from real‑world deployments, potentially accelerating the development of more adaptable robots that can be quickly re‑tasked for new environments.

Partnerships with major cloud and hardware providers like Nvidia and AWS suggest that Billy may leverage advanced and AI services, raising questions about data security, latency, and the cost structure of operating such robots at scale.

Interactive Mechanism

互動機制:它實際上是如何運作的

以互動方式探索這項發展背後的基礎技術。

Agent Lifecycle Stage:
1
User Intent & Planning: "Audit customer refund request #4092 and settle payment."
2
Tool Calling: Emits structured JSON call crm_get_transaction(id='4092').
3
Guardrail & Verification:🛡️ Paused: High-value action requires human operator sign-off.
4
Final Settlement: Refund recorded, email receipt dispatched, and audit log stored.
Core takeaway: An AI agent is not just a language model—it is a closed loop of planning, tool invocation, and environment feedback. Production systems require self-healing retries and strict human approval guardrails.
互動式概念檢查+10 Points
AI Agents Quiz

What most distinguishes an AI agent from a basic chatbot?

接下來看什麼

Key indicators to monitor include the outcomes of the CJ CheilJedang PoC and any subsequent contracts announced by Newbility with logistics or manufacturing firms. Pricing, leasing terms and the robot’s operational uptime under real‑world conditions remain undisclosed, so market reception will hinge on cost‑effectiveness compared with existing collaborative robots. Additionally, integration with cloud services from partners such as Nvidia and AWS may influence performance and data‑privacy considerations. Observers should watch for regulatory or safety certifications required for mobile manipulators operating in mixed human‑robot workplaces.

Results from the CJ CheilJedang proof‑of‑concept will be a primary barometer of Billy’s commercial viability in food manufacturing, a sector that demands high hygiene standards and reliable operation.

Future announcements regarding pricing models, leasing options, or subscription services will clarify the economic proposition for potential customers, especially small‑ and medium‑size enterprises.

Regulatory approvals for mobile manipulators in mixed human‑robot environments will be essential; any safety incidents or certification hurdles could affect rollout timelines.

The extent to which Newbility’s partners (Nvidia, AWS) provide specialized AI acceleration or cloud integration will influence performance benchmarks and may drive competitive responses from other robot manufacturers.

相關指引和測驗

人工智慧代理人工智慧模型解釋AI 的未來測試你所知道的—嘗試免費的人工智慧測驗在我們的詞彙表中尋找人工智慧術語關注 AI 模型發布追蹤器
覺得有用嗎?