长按二维码,关注官方公众号
使用微信扫一扫分享到朋友圈
使用微信扫一扫进入小程序分享活动
当具身智能从 “算法构想” 走向 “实体交互”，“大模型 + 多模态感知” 正成为技术破局的核心引擎：VLA 架构重构决策范式，VLM 让机器从 “看懂” 到 “理解”，而多传感器融合、跨模态对齐，则是打通 “感知 - 认知 - 行动” 闭环的关键。当前行业既迎来大模型下沉终端的机遇，也面临算力瓶颈、数据协同、端侧部署等现实挑战，“感知无界” 的技术突破，正定义具身智能的下一代产业格局。
从实验室的 “单点技术突破” 到产业端的 “全链路落地”，具身智能正卡在关键的 “衔接关口”：大模型的泛化能力难以适配机器人的实时感知需求，多模态融合的算法框架与硬件载体存在 “适配鸿沟”，端侧算力、数据协同的行业标准更是尚未统一。
在此背景下，盖世集团主办的“2026具身感知融合与多模态大模型创新研讨会”将于上海召开。活动聚焦 技术范式革命 、多模态感知融合、 VLA 架构、多模态感知、跨模态统一表征、算力与数据、端云协同、仿真进化、边缘推理、分布式训练、触觉 /视觉/力觉传感的落地实践；更有热点对话链接大模型、机器人、芯片等全产业链资源，汇聚产业头部玩家，打通具身智能从 “算法” 到 “实体” 的全链路。
诚挚邀请您拨冗出席本次研讨会，与国内外技术领袖、行业精英共探具身智能的未来路径，共同推动多模态感知与具身大模型的融合发展。期待与您相聚9月，携手迈向智能感知与实体行动无缝融合的新纪元！
Welcome to the The Symposium on Embodied Perception Fusion & Multimodal Large Model Innovation 2026, hosted by Gasgoo, to be held on September 15, 2026 in Shanghai.
As embodied AI moves from conceptual algorithms to real-world physical interaction, the convergence of foundation models and multimodal perception is reshaping the technological landscape. While VLA architectures and VLMs are redefining how machines perceive and reason, challenges remain in compute constraints, data coordination, edge deployment, and hardware-algorithm integration.
This symposium brings together the full ecosystem — from foundation models and robotics to chips and sensors — to address these challenges and chart the path forward. Topics include multimodal perception fusion, cross-modal representation, cloud-edge collaboration, simulation evolution, and practical sensor deployment.
We would be honored by your participation in this dialogue with global thought leaders and industry pioneers, as we work together to bridge the gap between intelligent perception and embodied action.
We look forward to seeing you in Shanghai.
演讲及参展请联系：冯女士 邮箱：conference@gasgoo.com，电话：021-39586681
For speaking engagements and exhibition inquiries, please contact: Ms. Feng，Email: conference@gasgoo.com，Tel: +86-21-39586681
具身智能从概念验证到规模化量产
盖世研究院
Embodied AI: From Proof of Concept to Mass Production
Gasgoo Research Institute
VLA重构机器人智能：从多模态理解迈向自主行动
智元/银河通用/智平方/星海图
VLA Reconstructs Robotic Intelligence: From Multimodal Understanding to Autonomous Action
AGIBOT/Galbot/AI Robotics/Galaxea
世界模型，通往通用智能的核心引擎
千寻智能/极佳视界/星动纪元
World Model: The Core Engine Toward General-Purpose Intelligence
Spirit AI/GigaAI/RobotEra
物理世界基座模型，构建机器人理解真实世界的底层能力
英伟达/光象科技
Physical World Foundation Model: Building the Underlying Capability for Robots to Understand the Real World
NVIDIA/Phi-Bot
茶歇&交流
Tea Break & Networking
从被动感知到主动探索，全域感知融合赋予机器人自主智能
帕西尼/跨维智能
From passive perception to active exploration, omnidimensional perceptual fusion endows robots with autonomous intelligence
PaXini Tech/DexForce
视觉-触觉融合突破机器人操作最后一厘米
优理奇/千觉/戴盟
Vision-tactile fusion breaks through the last centimeter of robotic manipulation
UniX AI/Xense Robotics/Daimon Robotics
感知硬件生态如何定义感知标准
大疆/禾赛
How does the perception hardware ecosystem define perception standards?
DJI/Hesai
热点对话：从技术突破到产业落地，具身智能的下一个增长点在哪里？
1. 未来三年，具身智能产业最大的突破机会在哪里？
2. 多模态感知、大模型、数据闭环，哪项能力最先形成产业竞争壁垒？
3. 哪些应用场景将率先实现规模化落地，并带动产业爆发？
4. 如何构建开放协同的具身智能产业生态，加速技术走向商业价值？
Hot Dialogue: From Technological Breakthroughs to Industrial Deployment — Where Is the Next Growth Engine for Embodied AI?
1. Where will the biggest breakthrough opportunities lie for the embodied AI industry over the next three years?
2. Among multimodal perception, foundation models, and data loop, which capability will become the foremost competitive barrier in the industry?
3. Which application scenarios will achieve large-scale deployment first and trigger an industry-wide takeoff?
4. How can we build an open and collaborative industrial ecosystem for embodied AI to accelerate the translation of technology into commercial value?
午餐&交流
Lunch & Networking
数据飞轮：开启持续进化时代
宇树/优必选
Robot Data Flywheel: Ushering in the Era of Continuous Evolution
Unitree/UBTECH
虚实融合数据体系：从“散装数据”到标准化工具链
光轮智能/无问智科
Virtual-Real Integrated Data Infrastructure — Transitioning from Fragmented Data to Standardized Toolchains
Lightwheel AI/WuwenAI
人类视频数据：低成本走通Scaling路线的核心路径
智在无界
Human Video Data: A Core Pathway to Achieving Scaling at Low Cost
BeingBeyond
大规模具身智能数据集的构建
星海图/智源研究院
Construction of Large-Scale Embodied AI Datasets
Galaxea/BAAI
Sim2Real 2.0：生成式仿真构建可泛化世界模型
苏度科技/诺亦腾机器人
Sim2Real 2.0: Generative Simulation for Generalizable World Models
Hillbot/Noitom Robotics
茶歇&交流
Tea Break & Networking
云端大脑与端侧小脑协同：规模化部署新架构
NVIDIA/华为/地平线
Cloud Brain and Edge Cerebellum Synergy: A New Architecture for Scalable Deployment
NVIDIA/Huawei/Horizon Robotics
端侧算力瓶颈突破：多模态融合推理的芯片级优化
爱芯元智/瑞芯微
Breaking Through the Edge Computing Bottleneck: Chip-Level Optimization for Multimodal Fusion Reasoning
Axera/Rockchip
云端AI芯片赋能具身智能实时决策
摩尔线程/寒武纪
Cloud AI Chips Empowering Real-Time Decision-Making for Embodied AI
Moore Thread/Cambricon
群体智能与多机协作：云边协同赋能规模化部署
北京人形机器人创新中心
Swarm Intelligence and Multi-Robot Collaboration: Cloud-Edge Synergy Enabling Large-Scale Deployment
Beijing Humanoid Robot Innovation Center
大会结束
End of Conference
具身智能从概念验证到规模化量产
盖世研究院
Embodied AI: From Proof of Concept to Mass Production
Gasgoo Research Institute
VLA重构机器人智能：从多模态理解迈向自主行动
智元/银河通用/智平方/星海图
VLA Reconstructs Robotic Intelligence: From Multimodal Understanding to Autonomous Action
AGIBOT/Galbot/AI Robotics/Galaxea
世界模型，通往通用智能的核心引擎
千寻智能/极佳视界/星动纪元
World Model: The Core Engine Toward General-Purpose Intelligence
Spirit AI/GigaAI/RobotEra
物理世界基座模型，构建机器人理解真实世界的底层能力
英伟达/光象科技
Physical World Foundation Model: Building the Underlying Capability for Robots to Understand the Real World
NVIDIA/Phi-Bot
茶歇&交流
Tea Break & Networking
从被动感知到主动探索，全域感知融合赋予机器人自主智能
帕西尼/跨维智能
From passive perception to active exploration, omnidimensional perceptual fusion endows robots with autonomous intelligence
PaXini Tech/DexForce
视觉-触觉融合突破机器人操作最后一厘米
优理奇/千觉/戴盟
Vision-tactile fusion breaks through the last centimeter of robotic manipulation
UniX AI/Xense Robotics/Daimon Robotics
感知硬件生态如何定义感知标准
大疆/禾赛
How does the perception hardware ecosystem define perception standards?
DJI/Hesai
热点对话：从技术突破到产业落地，具身智能的下一个增长点在哪里？
1. 未来三年，具身智能产业最大的突破机会在哪里？
2. 多模态感知、大模型、数据闭环，哪项能力最先形成产业竞争壁垒？
3. 哪些应用场景将率先实现规模化落地，并带动产业爆发？
4. 如何构建开放协同的具身智能产业生态，加速技术走向商业价值？
Hot Dialogue: From Technological Breakthroughs to Industrial Deployment — Where Is the Next Growth Engine for Embodied AI?
1. Where will the biggest breakthrough opportunities lie for the embodied AI industry over the next three years?
2. Among multimodal perception, foundation models, and data loop, which capability will become the foremost competitive barrier in the industry?
3. Which application scenarios will achieve large-scale deployment first and trigger an industry-wide takeoff?
4. How can we build an open and collaborative industrial ecosystem for embodied AI to accelerate the translation of technology into commercial value?
午餐&交流
Lunch & Networking
数据飞轮：开启持续进化时代
宇树/优必选
Robot Data Flywheel: Ushering in the Era of Continuous Evolution
Unitree/UBTECH
虚实融合数据体系：从“散装数据”到标准化工具链
光轮智能/无问智科
Virtual-Real Integrated Data Infrastructure — Transitioning from Fragmented Data to Standardized Toolchains
Lightwheel AI/WuwenAI
人类视频数据：低成本走通Scaling路线的核心路径
智在无界
Human Video Data: A Core Pathway to Achieving Scaling at Low Cost
BeingBeyond
大规模具身智能数据集的构建
星海图/智源研究院
Construction of Large-Scale Embodied AI Datasets
Galaxea/BAAI
Sim2Real 2.0：生成式仿真构建可泛化世界模型
苏度科技/诺亦腾机器人
Sim2Real 2.0: Generative Simulation for Generalizable World Models
Hillbot/Noitom Robotics
茶歇&交流
Tea Break & Networking
云端大脑与端侧小脑协同：规模化部署新架构
NVIDIA/华为/地平线
Cloud Brain and Edge Cerebellum Synergy: A New Architecture for Scalable Deployment
NVIDIA/Huawei/Horizon Robotics
端侧算力瓶颈突破：多模态融合推理的芯片级优化
爱芯元智/瑞芯微
Breaking Through the Edge Computing Bottleneck: Chip-Level Optimization for Multimodal Fusion Reasoning
Axera/Rockchip
云端AI芯片赋能具身智能实时决策
摩尔线程/寒武纪
Cloud AI Chips Empowering Real-Time Decision-Making for Embodied AI
Moore Thread/Cambricon
群体智能与多机协作：云边协同赋能规模化部署
北京人形机器人创新中心
Swarm Intelligence and Multi-Robot Collaboration: Cloud-Edge Synergy Enabling Large-Scale Deployment
Beijing Humanoid Robot Innovation Center
大会结束
End of Conference
Q: 是否支持开票？
A: 下单时填写开票信息，系统自动开具发票至用户邮箱。如下单时未选择开票，后续仍需开票或未收到系统邮件，可使用下单的手机号登录盖世汽车APP，在APP--我的--开票中选择对应订单，点击【申请开票】按钮，即可自主开票。
Q: 如何确认是否成功？
A：报名成功后系统会自动发送邮件及短信通知，并附专属个人签到码。也可以在下方找回电子票搜索框中输入报名时的手机或邮箱查询。
Q: 如何签到入场？
A：报名成功后，主办方签到系统已自动生成了您的专属个人<签到二维码>和<6位数签到码>，以短信及邮件的形式发送给您，请于会场签到处出示并签到，报名注册手机号也可用作签到。
Q：会议是否有视频直播？
A：会议实况会全程直播，关注【盖世智电产业观察】微信公众号，通过从【会议直播】菜单栏获取临期会议直播链接。
您还可以在盖世智电产业观察微信视频号、盖世汽车社区公众号、盖世汽车社区微信视频号、盖世汽车APP等平台关注直播。分段回放会后一周左右会发布在盖世汽车资讯网站、盖世智电产业观察微信视频号等渠道。
