Asia Emerging Robotics: Takeaways from Hesai's Tech Day andChina Humanoid Robot Half-Marathon humanoid robot half-marathon on April 19. In this note, we would like to share our view onthe technology and industry shifts reflected by the two events. +85221232622dien.wang@bernsteinsg.com +85221232631jay.huang@bernsteinsg.cor meaningful breakthrough was the full-color SPAD-SoC chip platform (Exhibit 1andExhibit 2), which integrates 2D image data and depth information directly at the chip-level.This represents a technological leap with thefollowing key advantages. +85221232666weibin.liang@bernsteinsg.com LiDAR point clouds, Hesai enables native and pixel-level fusion (Exhibit 3). This leads tobetter spatial andtemporal alignment,andultimatelyimproves accuracy,consistency,andreliability. We think it will be critical to Physical Al, as the improved perception may helpnarrow the sim-to-real gap and further elevate the role of simulation/synthetic data. - Greater flexibility. As the innovation starts at the chip-level, Hesai can develop full-colorLiDARs across a wide range of distance, field of view, and resolution. The rollout begins withthe ADAS-focused ETX series, scheduled forproduction in 2H2026. China Humanoid Robot Half-Marathon.The 21-kilometer race served as an extremestress test for the robots, and we highlight three key takeaways: -Rapid progress inrobot intelligence. Last year, most participants relied on remotecontrol; this year, ~40% of the participants (>100 teams) allowed their robots to runautonomously.The course includedflatsections,slopes,narrowpassages,and~20turns,demonstrating rapid improvement in robots'intelligence to handle generalizedenvironments in the real world. -Broad-based advancement of China's robotics industry.In Aug 2025,Unitree waslater,at this half-marathon,many brands'robots matched or even exceeded human runningspeeds, underscoring the industry's broad-based advancement (Exhibit 4). - Honor as a rising star. The biggest surprise was Honor, the Chinese smartphone maker,whose robots swept the top three places (Exhibit 5to Exhibit 7).Notably, Unitree alsoparticipated but did not place among the top finishers, suggesting that the technologicalgap between Unitree and other Chinese players is narrowing rapidly.This reinforces ourview since initiation (here)that it is still too earlyto identify the ultimate winners.With moretraditional giants entering the robotics industry, we believe they will rely on the highly-particularly for reducers. Against this backdrop, Shuanghuan is well positioned to be a keybeneficiary. EXHIBIT 3: Hesai: Spatial view of full-colored point clouds Source:Etown-halfmarathon, Bernstein analysis EXHIBIT 6:Honor's robot in the humanoid robot half-marathon INVESTMENTIMPLICATIONS WerateShuanghuan, Hesai and TuopuOutperform,Sanhua Market-Performand Leader Drive Underperform. References to"Bernstein"or the"Firm"in these disclosures relate to the following entities: Bernstein Institutional Services LLC (April 1, 2024 onwards), Sanford C. Bernstein & Co., LLC (pre April 1, 2024), Bernstein Autonomous LLP, BSG France S.A. (April 1,2024onwards),SanfordC.Bernstein(HongKong)Limited盛博香港有限公司,SanfordC.Bernstein(Canada)Limited,SanfordC. Bernstein (India) Private Limited (SEBI registration no. INH000006378), Sanford C. Bernstein (Singapore) Private Limited,Africa Technologies &Services toproduce Bernstein research under a Global Services Agreementinplacebetween BernsteinandSocieteGenerale otherwise, for purposes of these disclosures,references to Bernstein's“affiliates"relate to both SG and AB and their respectiveaffiliates. respectively,based on WACC of 8.3% and terminal growth rate of3.0%,respectively.They reflect ourlong term forecast of thecompany's existing core business and rising contribution from robotics.We use PEmethod as a reference, with our target priceimpliesa36xforwardPEatend-2026. We use DCF as the primary valuation method.We set target prices at CNY100.0based on WACC of 8.2% and terminal growthrateof3.o%,whichreflectsourlongtermforecastofthecompany'sexistingcorebusinessandrisingcontributionfromrobotics NingboTuopuGroup Co., Ltd. We use DCF as theprimary valuation method. We set target prices at CNY 75.0 based on WACC of 8.0% and terminal growth rateof 3.0%,which reflects our longtermforecast of the company's existing core business and rising contribution fromrobotics.WeusePEmethodasareference,withourtargetprice implies a32xforwardPEatend-2026 ZhejiangSanhuaIntelligentControls Co.,Ltd. We use DCF as the primary valuation method. We set target prices at CNY 39.0 and HKD 27.0 for its A-Share and H-Share,respectively, based on WAcC of 8.2% and terminal growth rate of 3.0%. They reflect our long term forecast of the company'sexisting core business and rising contribution from robotics.We use PEmethod as a reference, with ourtarget price implies a 34xforward PE at end-2026 Zhejiang Shuanghuan Driveline Co., Ltd. Weuse DCFastheprimary valuat