Woofun AI reports that Nianxiang Technology, a firm specializing in the industrialization of non-invasive neural interface technologies, has finalized an angel financing round totaling nearly 10 million yuan. The investment was spearheaded by Yongjun Xingmang, with additional participation from Pudong Venture Capital and Yicun Capital. These capital injections are designated specifically for accelerating product research and development, expanding the technical team, and constructing a localized surface electromyography (sEMG) dataset. This strategic funding arrival marks a pivotal moment for the company, which was established at the end of 2025 to bridge the gap between neural signal processing and consumer-grade human-computer interaction.

The company's inaugural product, Omniband, represents a distinct departure from traditional brain-computer interface methodologies that attempt to detect signals directly from the cerebral cortex. Instead, this wrist-worn device analyzes neuromuscular electrical signals within the wrist to interpret hand movement intentions and continuous dynamic gestures. By leveraging this approach, Omniband facilitates seamless interaction across a spectrum of devices including smartphones, computers, smart glasses, and smart home ecosystems. The underlying logic posits that capturing amplified muscle signals at the wrist offers a superior signal-to-noise ratio compared to raw brain signal decoding, thereby enabling a non-invasive solution that maintains high fidelity without surgical intervention.

Beyond immediate interaction capabilities, Nianxiang Technology identifies high-precision hand movement data as a foundational asset for the broader artificial intelligence sector. The organization aims to construct a large-scale sEMG dataset specifically tailored to local hand operation scenarios, addressing a critical gap in current global training resources. This dataset is intended to provide essential data support for the development of embodied intelligence, physical AI, and world model training. The strategic focus on data accumulation aligns with recent industry breakthroughs, specifically research published by Meta Platforms in Nature, which demonstrated that the Scaling Law effect allows for cross-user, calibration-free recognition through sufficient data volume. This scientific validation, emerging in 2025, effectively unlocked the industrialization path for neural wristbands that previously struggled with generalization issues and lightweight design constraints.

Dr. Wang Yi, the founder of Nianxiang Technology, brings extensive academic and industry credentials to this venture, having conducted doctoral research on brain-computer interfaces at the University of Auckland in New Zealand. Currently serving as the vice chairman of the National Brain-Computer Interface Industry Alliance and a member of the Shanghai Magnolia Talent Program, Wang Yi previously held the role of chief scientist at Yingmai Medical and R&D director at Zhiyuan Robotics. His professional history encompasses deep engagement in the research and development of invasive, non-invasive, and interventional brain-computer interfaces, alongside sEMG neural interface technologies. Wang Yi's core philosophy centers on the interpretation of human intention, yet he explicitly rejects invasive solutions or medical-grade devices for mass adoption due to their lack of user acceptability and frequency of use. He argues that for brain-computer interfaces to truly benefit ordinary people, the technology must be solid, wearable, and integrated into daily life, a realization that drove the decision to pursue the neural interface wristband direction despite historical challenges.

Woofun AI data shows that the technical hurdles facing sEMG devices, such as motion artifacts, signal drift, and environmental noise, are being addressed through a systematic triad of hardware, algorithmic, and model-level innovations. In terms of hardware architecture, the team employs differential electrodes and optimized structural designs to reduce common-mode noise and enhance wearing stability, thereby minimizing signal drift caused by positional changes or limb movements. On the algorithmic front, custom filtering and signal separation techniques have been developed to eliminate interference stemming from motion artifacts and skin sweat. At the model level, the integration of multimodal data supplementation and cross-validation mechanisms enables AI models to learn comprehensive signal characteristics, significantly enhancing robustness across diverse users and action states. These combined efforts allow the current Omniband prototype to continuously estimate the dynamic angles of all 20 joints in the hand, a capability that far exceeds the basic metrics recorded by traditional fitness trackers.

The differentiation between Omniband and conventional fitness sensors lies in the depth of data interpretation and the nature of the output. While standard trackers record elementary data points such as step counts, heart rate, and exercise duration, Omniband captures the user's movement intentions directly, analyzing joint angles and muscle activity levels to identify fine hand movements and continuous dynamic gestures. Through the standard HID Bluetooth protocol, the device enables remote control of various electronic systems and facilitates air-based handwriting, effectively creating an "invisible" keyboard and mouse that removes the physical limitations of traditional input devices.

Furthermore, the high-precision data collected serves as a critical input for embodied intelligence and physical AI development, with plans to incorporate additional sensors to expand application ranges in the future. The product currently resides in the engineering prototype stage, with the team prioritizing the refinement of calibration-free performance and cross-user generalization.

Regarding commercialization and market entry, Nianxiang Technology has adopted a phased strategy that prioritizes the B-side market before expanding to consumer segments. The initial focus involves providing customized interaction solutions, embodied data collection services, and SDK licensing to universities and large enterprises. This approach allows the company to validate its technical solutions in real-world scenarios while continuously accumulating diverse data to improve model capabilities. As the product matures and the dataset expands, the company plans to gradually extend its offerings to geek communities, technology enthusiasts, and eventually the broader consumer market. Although the product is not yet fully "out of the box ready", new users can currently complete a quick 30–60-second calibration process and begin usage after performing a few basic actions, indicating a relatively low learning curve. The ultimate goal is to lower the usage threshold to a level acceptable for mass-market consumer products through continued improvements in calibration-free recognition and cross-user accuracy.

The broader implications of Nianxiang Technology's progress extend beyond a single product launch, signaling a potential shift in how human-computer interaction is conceptualized and executed. By leveraging the Scaling Law effect and focusing on the construction of a local sEMG dataset that rivals ImageNet in scope, the company is addressing the fundamental generalization problems that have long plagued the industry. The ability to collect multi-dimensional data including hand posture, muscle activity, and object interaction through a non-invasive wristband suggests a viable path toward ubiquitous neural interfaces. As the team continues to recruit volunteers and open a developer platform to enhance model capabilities, the trajectory points toward a future where physical input devices become obsolete, replaced by intuitive, gesture-based control systems. This marks a significant evolution in the field, moving from experimental medical applications to practical, everyday tools for digital interaction.