Beijing time, October 10, TIME magazine unveiled its 2025 Best Inventions list, spotlighting 300 groundbreaking products, with DeepSeek’s R1 inference model and Unitree Technology’s R1 humanoid robot standing out as beacons of Chinese innovation. At Jeenoce, a leader in digital transformation for electronics manufacturing, we extend heartfelt congratulations to these trailblazers. Their achievements signal a global shift driven by the dual engines of digitalization and intelligence—our very “wings” to soar in the smart factory era.
DeepSeek’s R1, hailed as “AI’s resounding first shot,” debuted in January 2025 as an open-source model with 32B and 70B parameters, rivaling OpenAI’s o1 in math and coding benchmarks while using fewer computational resources. Meanwhile, Unitree’s R1 humanoid robot, with 26 joints (compared to human’s ~300), performs complex tasks like boxing, running, and somersaults. Weighing only 25kg, standing 1.2m tall, and equipped with AI for voice and image processing, R1 starts at $5,900, targeting research and education. These innovations mirror the core challenges and opportunities in modern manufacturing: how to make production “brains” as swift as DeepSeek and “nervous systems” as precise as Unitree?
DeepSeek R1’s success lies in its cold-start data integration and reinforcement learning (RL), enabling advanced reasoning with limited resources—a principle echoed in Jeenoce’s Manufacturing Execution System (MES). Our MES acts as an AI-like core, collecting real-time production data from raw material intake to finished goods dispatch. Through edge computing, it analyzes sensor signals for predictive maintenance and dynamic scheduling. For instance, in complex electronics assembly, MES mimics DeepSeek’s “efficient inference,” identifying bottlenecks early, reducing human intervention, and boosting decision-making speed by over 50%.
Our digital twin technology further parallels Unitree’s simulation-based robot training. By creating high-fidelity 3D virtual factory models synced with real-time data, we simulate “what-if” scenarios, such as optimizing SMT machine paths or predicting equipment failure rates. Like Unitree R1’s 24-26 degrees of freedom (DOF) training environment, our digital twins iterate thousands of scenarios virtually, cutting physical trial-and-error costs. Applied in client factories, this technology mitigates supply chain volatility, slashing pilot production cycles.
Unitree R1’s 26 joints and AI-driven actuators transform abstract commands into fluid motions, a feat mirrored by Jeenoce’s custom automation equipment. In electronics manufacturing, our non-standard equipment—like multi-axis robotic arms and vision-guided systems—serves as the “nervous system,” converting MES digital signals into millimeter-precision physical actions. For high-density PCB soldering, our equipment integrates servo motors and force-feedback sensors, adapting to variable workpieces (e.g., diverse chip packages) with zero-defect execution, akin to Unitree’s agility.
These “neural” components support modular upgrades via EtherCAT bus, achieving sub-millisecond synchronization, surpassing traditional PLC systems. This enables “flexible manufacturing,” allowing rapid reconfiguration for diverse scenarios, from consumer electronics to automotive components, in the spirit of Unitree’s versatile design.
To illustrate, we share a desensitized architecture diagram of Jienuosi’s Digital Factory System (DFS). It depicts a closed-loop structure: data collection (sensors and IoT), analysis (MES and AI engines), and execution (automation equipment). The user interface features real-time KPI dashboards, enabling one-click historical trend analysis.
The list also highlights Huawei’s Pura 80 Ultra, BYD’s Seagull, and Apple’s AirPods Pro 3, signaling intelligence permeating electronics and transportation. Jienuosi Electronics will harness these dual engines, collaborating with partners to shape a digital future. We welcome industry peers to explore AI and automation synergy with us!
Message
返回列表