When embodied intelligence emerges as one of the most prominent topics at this WAIC, the robotics industry is entering a new phase of development.

刚刚落幕的2026世界人工智能大会（WAIC），汇聚了全球人工智能领域的前沿技术与创新成果。借助大模型以及智能终端方面的进展，从机器人到产业应用，具身智能无疑成为了现场最热的关键词之一。

陈士凯科技创始人兼CEO陈士凯受邀出席「具身人形机器人落地与核心零部件发展论坛」发表主题演讲，并参与圆桌论坛，与产业伙伴共同探讨具身智能时代机器人核心技术的发展方向。

Focused on embodied intelligence, the underlying capabilities of robots are undergoing a new upgrade.

在以《构建具身智能时代多模态空间感知和自主行动平台》为主题的演讲中，陈士凯围绕机器人空间感知以及移动平台的发展趋势展开讨论，分享了思岚科技对于具身智能底层能力建设所进行的思考。

Current wheel and legged embodiments have emerged as the two primary technical routes for embodied intelligence robots, with wheeled robots that combine mobility efficiency and application flexibility remaining a key direction for industrialization and deployment at this stage. Traditional wheeled robots nonetheless continue to confront multiple challenges:

1: The limited degrees of freedom in motion make it difficult for the robot to adapt to narrow and complex spaces

2: 依赖于二维激光导航，难以满足三维环境感知需求

Traditional SLAM algorithms are also increasingly difficult to support the requirements for environmental understanding and intelligent decision-making in the era of AI large models.

With years of accumulated foundational technology, Shenzhen Technology has established a full-stack technical capability encompassing laser radar, multimodal perception algorithms, and intelligent mobile chassis. To meet the high-level deployment requirements of humanoid robots, Shenzhen Technology has newly launched the Poseidon omnidirectional mobile robot chassis, which serves as the core technology deployment benchmark in this speech.

✅ 全向运动，突破传统移动局限

Adopting the industry-leading four-wheel four-steer and four-wheel eight-drive architecture, the chassis supports lateral translation, in-place rotation, diagonal driving, and parking modes, thereby completely resolving the traditional chassis issues of rigid motion and cumbersome steering while meeting the flexible mobility needs for adaptation to narrow spaces and complex working conditions.

四轮独立转向

✅ 多传感器融合，适配人机协作场景

标配前双激光雷达可选配3D多线雷达双深度视觉与超声波传感器构建360°无死角感知体系结合自研SLAM算法实现三维全场景感知精准避障适应高阶人机协作复杂环境自主作业需求。

✅ SLAMWARE NX 新一代自学习定位引擎

配套专属迭代的Slamware NX具身智能版导航系统，完成了从2D激光SLAM到视觉激光融合3D SLAM的全面升级，其中在线自主学习和修正使得环境变化率超过80%时依旧能提供长期稳定的可靠定位。

✅ 全形态灵活适配，极速开发落地

全模块化架构承载稳定并适应灵活，支持裸底盘、定制外壳以及整机上装等形态，兼容折叠、升降、丝杆等多种上装结构，提供标准和超低重心双版本，最低安装面仅65mm，可以满足各类个性化定制需求。

依托多年SLAM导航技术积淀，集成建图、路径规划、智能避障以及多机调度全栈能力，原生兼容ROS/ROS2并配备成熟SDK，可将整机开发周期压缩至1个月，大幅降低研发门槛。

除了移动能力之外，机器人理解环境同样是具身智能的重要基础。

针对机器人从二维几何地图迈向三维语义地图的发展趋势，陈士凯还介绍了思岚科技Aurora系列3D定位建图解决方案。

Aurora series, which integrates binocular vision, deep learning algorithms, and SLAM technology, enables high-precision positioning, real-time three-dimensional mapping, and semantic perception, allowing robots to upgrade from perceiving space to understanding space and thereby further enhancing their autonomous perception and decision-making capabilities in complex dynamic environments.

陈士凯表示：

未来具身智能的发展并非单一产品的竞争，而是空间感知定位导航以及移动平台等底层能力的系统融合。

| 巅峰对话，共探国产机器人自主可控之路

陈士凯还参与了论坛圆桌对话，与来自产业链上下游的嘉宾围绕《国产人形机器人在核心零部件如何实现：国产自主可控与规模化应用？》展开了深入交流。

对于主持人提出的聚焦国产人形机器人规模化落地与出海围绕供应链自主可控核心部件卡脖子及国产替代问题思岚科技陈士凯也做出了解答。

Laser radar technology faced early monopolization by overseas entities, which served as the fundamental reason for the establishment of Shenzhen Technology and its commitment to self-developed mass production. With accumulated technological progress, breakthroughs have now been achieved, enabling the sensor sector to break free from the neck-strangling dilemma entirely. Domestic substitution has matured, becoming both widely applied in the industry and fully reliable.

AI and embodied intelligence are experiencing sustained positive developments in the industry, as seen in Shenzhen Technology's offerings. Both the general-purpose chassis designed for embodied intelligence robots and the 3D localization mapping solutions tailored for complex environments and embodied intelligence applications have achieved a domesticization rate of over 99 percent.

同时，陈士凯还指出，真正长期制约行业的关键在于。

Standard discourse rights deficiency stands as the core challenge that the future requires us to confront for the long term.

这也是思岚科技持续深耕国家团体标准编制工作的核心原因。多年来思岚科技积极履行行业责任，深度参与多项智能机器人核心零部件、算法测评与场景应用类团体标准制定，对行业标准空白进行填补，对技术与测评规范进行统一，以标准化建设引领产业规范化、高质量发展。

在讨论“依托高性价比优势的中国人形机器人将在哪个时间节点迎来大规模市场放量？以及哪类人形机器人、适配什么场景、承接何种作业会率先实现市场爆发？”时，陈士凯对国内具身智能机器人产业化前景更为乐观，他认为：

国内拥有独特的落地优势，产业场景对新技术包容度更高，试错意愿更强，因此技术落地和迭代速度显著领先海外。

行业产业拐点已然临近，过往扫地、送餐机器人的爆发依托底层技术成熟，如今国内具身智能、国产大模型高速迭代，软硬件配套持续完善，已具备规模化落地的基础条件。

产业节奏层面： 具身智能机器人全民普及尚需3-5年，但细分场景规模化爆发将于明年中下旬开启。今年下半年至明年，设备感知、移动及大模型方案将逐步成熟，可覆盖大部分场景需求，叠加后续算力升级、2027年硬件成本大幅下降，产业落地门槛将持续走低。

Scene landing aspect: Flexible manufacturing and small-to-medium factory tail-end scenarios will see priority in scaled deployment, proving superior to traditional high-precision industrial scenes. These scenarios feature high tolerance for errors, emphasis on flexibility, and no requirement for extreme precision, thereby highly aligning with the embodied intelligence advantages of high generalization and low cost. This makes them the core track for the initial scaled landing of humanoid robots.

思岚科技始终坚持做好机器人底层能力。

近年来，人工智能正加速从数字世界走向物理世界，而机器人正成为这一进程的重要载体。

In more complex application scenarios, the need arises for stable and reliable positioning and navigation capabilities, more precise spatial perception abilities, more open and easy-to-use intelligent robot mobile chassis, and faster-deployed solutions.

未来，思岚科技将继续与产业伙伴携手合作，不断探索机器人核心技术的创新，共同推动具身智能走向更多真实应用场景，让机器人真正成为连接人工智能与现实世界的重要桥梁。

WAIC 2026, though it has already concluded, marks the beginning of an era of embodied intelligence that is accelerating. Shenzhen Technology will also continue to drive technological innovation and, together with industry partners, embark on the next chapter of the robotics industry.

来源：WAIC 2026｜思岚科技亮相世界人工智能大会，聚焦具身智能底层能力建设-资讯中国 | 资讯中国