August 1, 2026.

This morning, China's first batch of national standards for brain-computer interfaces went into effect — GB/T 47023-2026 Reference Architecture and GB/T 47127-2026 Multimodal Data Format. On July 28, at the Yicai Tech Innovation Summit in Shanghai, industry leaders from BrainTalker, Stairmed, Xiangyu Medical, and other BCI pioneers gathered to debate the same question:

After decades of lab demos, is BCI finally ready for the market?

The consensus: yes. But the real question isn't whether — it's where, how, and who pays.

this week, I had the privilege of serving as a judge for eight non-invasive BCI startups at an iFLYTEK innovation competition. Without disclosing any specific projects or proprietary details, I can say this: the patterns I observed — where teams are betting, where capital is flowing, where talent is converging — confirmed and deepened the bigger-picture trends I've been tracking. Combined with ongoing industry research, here's my synthesis across five dimensions: China's industrial geography, global competitive positioning, commercialization pathways, active capital, and cross-border strategy.

I. Five Commercialization Bets — and Where the Money Is (or Isn't)

Bet 1: Implantable Clinical BCI — "The Main Event"

BrainTalker has entered GCP-registered clinical trials through a consortium of 15 tertiary hospitals led by Huashan Hospital. Stairmed has completed 50+ investigator-initiated implants and its 256-channel fully implantable system has completed clinical implantation, with a large-scale multi-center registration trial planned this year. These are the "serious BCI" plays — regulatory moats, decade-long R&D timelines, and the largest potential addressable market. Think Neuralink, Synchron, BrainGate.

My take: This is where the big outcomes will be, but the timeline is measured in years, not quarters. Three-year tipping point for first devices; five-to-eight years for broader crossover. Marathon, not sprint.

Bet 2: Non-Invasive Rehabilitation — "The Quiet Engine"

Xiangyu Medical (SHA: 688626) has quietly put its BCI rehab devices into over 500 hospitals, aiming for 1,000+ by end of 2026. It holds NMPA registrations for EEG devices — a regulatory moat many startups can only dream of.

But here's the truth: as of end-2025, BCI products still haven't reached "scaled sales" even for a publicly traded company with 500 hospital installs. That tells you how early this market still is.

My take: Non-invasive rehab is the most bankable BCI commercialization thesis right now. The question is how fast "deployed" converts to "paid for."

Bet 3: Clinical Vertical BCIs — "The Underrated Gems"

Epilepsy closed-loop management, ASD pediatric assessment, and similar clinical-vertical plays are where I see the most overlooked alpha. They have: (1) a clear clinical stakeholder, (2) a measurable endpoint, and (3) a natural regulatory pathway. They can price like medical devices, not SaaS subscriptions. Risk: boutique markets.

Bet 4: Consumer EEG Wellness — "The Crowded Middle"

A growing number of teams are entering the consumer EEG space with the same fundamental stack: dry-electrode EEG + deep learning classification + SaaS subscription. This is a structural warning, not a feature contest. Commodity input, not differentiated product. Low barrier to entry, zero barrier to pricing power.

Bet 5: BCI-Control Platforms — "Visionary, but Early"

Infrastructure plays that require platform-level adoption and significant capital. Valuable long-term direction, but favors large platforms over early-stage startups.

II. China's BCI Industrial Geography

China's BCI industry is no longer a single-axis narrative. Multiple cities now have dedicated policies, dedicated funds, and dedicated industrial parks — and they're competing as much as they're collaborating.

Guangdong's June 26 Action Plan is the first provincial-level systematic BCI deployment framework. Hubei published the first BCI pricing items in 2025; Sichuan followed with 82 categories. But most BCI services remain Category C (fully self-paid) — priced, but not yet reimbursed. For anyone modeling revenue: "priced" ≠ "paid for."

III. The Reimbursement Question

Three models emerging: medical insurance, enterprise payment, direct-to-consumer. The most concrete data point: Boruikang's NEO system — world's first approved invasive BCI Class III device (March 2026). Within one week: Shanghai medical consumables catalog (procurement clearance). May: "Huhuibao" added BCI coverage (30% / ¥150K). June: ¥2.5B STAR Market IPO filed. July 13: first commercial implant at Huashan Hospital. Most complete "lab→approval→pricing→reimbursement→capital markets" chain — but NEO has zero revenue in the reporting period. Approved and profitable are still years apart.

My core conviction: Hospital procurement → insurance framework → employer wellness → consumer DTC. Consumer-first BCI without a clinical anchor is playing the wrong end of the funnel.

IV. The Capital Landscape

• Corporate VCs: Botuo Biotech 4 BCI investments in 2 years. Dinake + USTC Joint Lab.
• Big tech arrived: BrainCo ~¥2B B+ round. Stairmed ¥500M led by Alibaba, Tencent topping up. Gestalt Tech ¥570M in six months.
• Government as LP: Shanghai state funds deployed across multiple BCI companies.

H1 2026 BCI financing exceeded ¥7 billion — more than double full-year 2025. When capital floods a sector with multi-year commercial timelines: valuations detach from revenue, teams optimize for rounds over milestones, and the correction punishes everyone. This pattern has played out in CAR-T and digital therapeutics. It's happening now in BCI.

V. Cross-Border Strategy

1. Clinical collaboration: Dual-region filings (NMPA + FDA/CE) in motion.
2. Manufacturing supply: China's BCI hardware ecosystem becoming a global supplier base.
3. Emerging markets: Non-invasive BCI cost-competitive for SE Asia, Middle East, Africa.

My Five Questions

1. Who pays, and when? Name the buyer and their procurement cycle.
2. Is your moat regulatory or algorithmic? Regulatory moats last years. Algorithmic moats last months.
3. Product or feature? Features get acquired. Products build companies.
4. Capital efficiency? Smallest check to prove revenue?
5. Global differentiator? What stops a US/EU competitor tomorrow?

The BCI industry is where oncology immunotherapy was in 2014 — credible science, early signals, too many companies chasing too few indications, and a handful of compounders quietly building defensible positions.

2026年8月1日。

今天，中国首批脑机接口国家标准正式实施——GB/T 47023-2026《参考架构》和GB/T 47127-2026《多模态数据格式》。三天前，在第一财经科技创新大会上，脑虎科技、阶梯医疗、翔宇医疗等BCI领军企业齐聚上海，讨论同一个问题：

实验室做了几十年，脑机接口到底能不能赚钱了？

共识是：能。但真正的问题不是"能不能"，而是——在哪赚？怎么赚？谁来买单？

本周，我作为评审参与了科大讯飞的创业大赛，评审了八家非侵入式脑机接口赛道创业项目。在不透露具体项目创新点和细节的前提下，我可以分享一个观察：这些团队在商业化方向上的分化，与我长期追踪的产业趋势高度吻合。结合持续行业研究，我从五个维度梳理判断：中国产业布局、全球竞争力、商业化路径、资本格局、跨境出海。

一、五条商业化路线——钱在哪儿、陷阱在哪儿

路线一：植入式临床BCI——「主战场」

脑虎科技已通过华山医院牵头、联合15家三甲医院启动GCP注册临床试验。阶梯医疗则在推进一系列IIT病例，其256通道全植入系统已完成临床植入，计划年内启动大规模多中心注册临床。对标：Neuralink、Synchron、BrainGate。

我的判断：最大的结果一定出在这条线上，但时间尺度是年不是季度。三年内首批器械商业化井喷；五到八年向医疗消费领域溢出。马拉松，不是百米。

路线二：非侵入式康复——「沉默的印钞机」

翔宇医疗（科创板688626）已把BCI康复设备铺进500多家医院，今年目标超千家。手握NMPA脑电图机注册证。但截至2025年底，BCI产品仍"尚未实现规模化销售"。一家上市公司、500家装机，BCI还没成为实质性收入线。这告诉你这个市场有多早期。

我的判断：非侵入式康复是当下最可押注的BCI商业化路径。问题是"装机"多快能变成"付费"。

路线三：临床垂直BCI——「被低估的珍珠」

癫痫闭环管理、ASD儿童评估——有明确临床买单方、可量化终点指标、天然监管通路。可以定溢价，而不是卷SaaS订阅价。风险：精品市场。

路线四：消费级EEG健康——「拥挤的中间地带」

越来越多的团队涌入，技术架构高度一致：干电极EEG+深度学习+SaaS订阅。这是同质化的生产要素，不是差异化的产品。准入门槛低，定价权门槛更低。

路线五：BCI通用控制平台——「远见但太早」

方向正确，但需要平台级采纳和大量资本。天然偏向大平台和资金充裕的研究联合体。

二、中国BCI产业地理

中国的BCI产业已经不是单轴叙事。多个城市已出台专项政策、专项基金和专项产业园——它们之间的竞合关系正在定义产业格局。

广东6月26日行动计划是首个省级系统性BCI部署框架。此前湖北已率先出台BCI医疗服务价格，四川跟进整合82项。但当前BCI服务多为丙类（完全自付）——定了价但还没报销。对算收入时间表的人来说，"定价"和"报销"的区别就是一切。

三、支付：谁来买单？

最具体的数据点：2026年3月，博睿康NEO系统获批全球首款侵入式BCI三类医疗器械。获批10天内即挂网上海医用耗材目录，打通医院采购通道。5月沪惠保首次增加BCI手术耗材保障（30%/15万上限）。6月科创板IPO获受理，拟募资25亿。7月13日首例商业化植入手术在华山医院完成。BCI史上最完整的"实验室→获批→定价→报销→资本市场"链条——但NEO报告期内收入为零。从获批到盈利，仍然以年为单位。

我的核心判断：医院采购→医保框架→企业健康→消费者DTC。跳过临床锚点直接押注BCI消费化，赌错了漏斗方向。

四、资本版图

• 上市公司领跑：博拓生物两年四度落子。狄耐克+中国科大BCI联合实验室。
• 大厂进场：强脑科技~20亿B+轮。阶梯医疗阿里领投5亿，腾讯作为老股东加注。格式塔科技半年5.7亿。
• 政府成为LP：上海国资全面布局BCI企业。

2026年上半年国内BCI融资超70亿，是2025全年的一倍多。资本涌入年计赛道的结果：估值脱离收入、团队优化为下一轮而非里程碑、修正惩罚所有人。这个模式在CAR-T和数字疗法上都发生过，现在正在BCI上重演。

五、跨境出海

1. 临床合作：双区域注册（NMPA+FDA/CE）已在植入式领域启动。
2. 制造输出：中国BCI硬件生态正在成为全球供应商基础。
3. 新兴市场：非侵入式BCI在东南亚、中东、非洲的成本优势明显。

我给这个行业的五个追问

1. 谁来买单？什么时候买？说出买方名字和采购周期。
2. 监管护城河还是算法护城河？注册证管几年，算法管几个月。
3. 产品还是功能？功能被收购，产品造公司。
4. 资金效率？花最少的钱跑通第一个付费闭环。
5. 全球差异点？海外BCI公司明天进你的市场，你拿什么挡？

脑机接口现在的状态，很像2014年的肿瘤免疫治疗——科学可信、早期商业化信号清晰、太多公司追太少已获批的适应症、真正能复利的几家在悄悄挖护城河。