Summary
Recently, Ma Daolin, founder and CEO of Xense Robotics, published an op-ed titled "Why tactile intelligence may become…
Recently, Ma Daolin, founder and CEO of Xense Robotics, published an op-ed titled "Why tactile intelligence may become the next infrastructure layer for physical AI" at the World Economic Forum, sharing his views on the next-stage technological focus of Physical AI, the value of tactile intelligence, and how robots can truly move into the real world.
When robots truly enter factories, warehouses, hospitals and homes, is merely "seeing" the world enough?
The article points out that as AI moves from the digital space into the physical world, the core question facing robots is no longer just "can they understand the world" but "can they reliably interact with the world." Therefore, the next stage of Physical AI may not depend only on larger-scale models, but on whether robots can acquire richer and more complete physical perception capabilities. Among these, the feedback provided by touch — contact, force, friction, deformation and material state — is becoming an important layer connecting intelligence with the physical world.
Since its founding, the Xense team has been deeply committed to the tactile domain. It not only leads in theoretical exploration over the long term, but also combines depth and foresight in technology commercialization. Today, Xense has built the full-stack "hardware-data-model" ecosystem required for physical interaction, and is also the only industry pioneer in China with proven real-world deployment capabilities. In the future, Xense will accelerate the commercialization of Physical AI through breakthroughs in the tactile domain.
About Xense Robotics
Xense Robotics Technology (Shanghai) Co., Ltd. was founded in May 2024 by Ma Daolin, a professor at Shanghai Jiao Tong University and winner of the only Best Paper Award at ICRA 2021. The company is a high-tech enterprise oriented toward the era of physical intelligence, dedicated to intelligent robots' perception of and autonomous interaction with the real physical world.
Xense has built a complete technology system covering hardware, data, models and scenario deployment, and has independently developed multimodal tactile sensors, the XTac tactile data collection system and the VTLA multimodal foundation model. The VTLA model integrates vision, touch, language and action information, aiming to enhance robots' perception, understanding and manipulation capabilities in real physical interaction, further strengthening the company's technological moat in tactile intelligence. Leveraging the closed loop of "hardware-data-model-scenario," Xense drives the deployment of tactile perception from underlying hardware and real interaction data to embodied model training and application deployment.
To date, Xense has served more than 300 leading industry companies and top research institutions, and collaborates with Li Auto, CATL, L'Oréal, AgiBot, TARS Robotics and others, with strategic investments from Hillhouse Ventures, Oriza Holdings, Li Auto and more. Going forward, Xense will continue to build the "tactile foundation" for the embodied intelligence era, advancing robots from "seeing the world" toward "understanding and interacting with the world."
