![]() |
市場調查報告書
商品編碼
2109329
OEM廠商和ADAS一級供應商在智慧駕駛海外擴張方面的突破性策略(2026年)Report on Breakthrough Strategies of OEMs and ADAS Tier 1 Suppliers for Overseas Layout of Intelligent Driving, 2026 |
||||||
監理突破、在地化體系建立、OEM廠商競相爭奪NOA佈局:2026年中國智慧駕駛海外佈局整體趨勢
海外自動駕駛部署情況調查-國內市場現有市場佔有率的競爭已達到頂峰,全球化對於自動駕駛的發展至關重要。
中國乘用車市場已徹底結束漸進式擴張週期,進入庫存競爭階段。 2020年至2025年,中國乘用車銷售年均複合成長率預估僅2.34%,汽油車市佔率持續萎縮,產業內價格競爭成為常態,內向型趨勢加劇。在日益飽和的國內市場中,汽車製造商已無法同時支撐大規模產能擴張和長期高水準的智慧技術投資,產業成長的限制日益凸顯。
與國內市場成長乏力形成鮮明對比的是,中國汽車出口經歷了爆炸性成長,從一項輔助業務發展成為第二大支柱產業。出口量從2020年的不足100萬輛飆升至2025年的709.8萬輛,短短六年間成長了七倍。同時,出口滲透率從3.9%躍升至20.6%,出口業務的年均複合成長率高達48.14%。海外市場已成為支撐中國汽車製造商擴張和發展的核心驅動力。
從全球市場展望來看,未來3-5年將是全球汽車製造商格局重組、產能和品牌層級鞏固的關鍵時期。根據長安汽車預測,到2030年,全球汽車市場規模預計將達到1億輛,新能源汽車滲透率將達到45%。產業內的「馬太效應」將進一步加劇,競爭門檻也將隨之提高,層級結構將會更加明確。對企業而言,年銷量達到300萬至350萬輛是生存的最低標準,500萬至700萬輛則表明企業業績穩健,而達到800萬至1000萬輛則有助於企業在全球範圍內建立競爭優勢。
企業僅靠國內市場已無法生存。全球海外擴張已從「附加選項」演變為「汽車製造商在下一階段競爭中的核心成長引擎」。各大汽車製造商正在徹底轉變發展邏輯,從「聚焦中國市場並向海外擴張」轉向加速轉型,成為擁有全球研發、製造、銷售管道和品牌能力的綜合性跨國集團。
許多國內汽車製造商正在將其策略重點從「專注於國內市場+拓展海外」轉向「加速轉型為擁有全球研發、製造、銷售管道和品牌體系的全球汽車集團」(2026-2030年)。
比亞迪的目標是成為全球規模最大的汽車製造商。到2026年,比亞迪計劃在全球整體銷售超過500萬輛汽車(其中海外銷量超過150萬輛),到2027年,銷量超過600萬輛,力爭全球市場佔有率超過10%。到2030年,比亞迪計劃鞏固並進一步加強其作為全球第一大新能源汽車製造商的領先地位。
2026年,東風汽車公佈了「揚帆起航」全球化策略,投資數千億元人民幣,推動全價值鏈海外擴張,遵循「全球產品部署—全球產能部署—全球產業部署」三階段路徑。到2030年,公司目標是全球整體銷售量達到500萬輛,其中新能源汽車佔比超過70%,海外銷售佔40%。
長安汽車正在升級其「大洋計劃2.0」,推動海外業務從「產品貿易」向「產業紮根」的深度轉型,力求從汽車出口商向全球汽車集團實現戰略轉型。公司目標是到2030年全球銷售突破500萬輛,躋身全球十大汽車製造商。公司設定了150萬輛的海外銷售目標,並雄心勃勃地設定了180萬輛的目標,計畫建造80萬輛的海外產能體系。
國內自動駕駛市場競爭日益激烈,加速了該技術在全球的傳播。
中國的自動駕駛產業正在快速擴張,市場正迅速從逐步普及階段過渡到現有市場的激烈競爭階段。這種日益激烈的競爭是推動汽車製造商拓展海外市場的主要動力之一。目前,自動駕駛功能已成為乘用車的標配,而非選配。
數據顯示市場結構發生了根本性轉變。配備導航就緒(NL)技術的車輛佔比從2022年的49.6%下降至2026年第一季的21.8%。同時,L2級以上自動駕駛車輛的滲透率從34.7%成長至69.8%。基礎L2 ADAS仍是市場基礎,而高速公路NOA和都市區NOA等高階功能自2024年起成為唯一高成長的細分市場。
目前,國內自動駕駛市場在各價格區間均已趨於飽和。在售價超過20萬元的高階車型中,L2級及以上自動駕駛功能的普及率已達80%至90%,而售價在10萬至20萬元之間的主流車型中,普及率也達到了60%至70%。另一方面,先進技術的應用正在穩步推進。在售價低於10萬元的入門車型中,自動駕駛功能的普及率從2022年的1.6%成長至2026年第一季的16.7%。國內自動駕駛市場正面臨成長潛力收窄和利潤率承壓的局面。技術和能力的海外產出已成為產業發展的必然趨勢。
主要障礙在於,中國的自動駕駛產業在技術和產業鏈方面都擁有世代優勢,並擁有全球規模的生產能力。
與歐美、日本、韓國等市場相比,中國自動駕駛產業在技術架構、工程技術和整個供應鏈三個方面已建立起世代優勢。這使其成為國內少數幾個擁有高海外市場競爭力並具備技術出口能力的核心產業之一。
中國在高級自動駕駛技術的廣泛應用和迭代方面領先世界。中國自動駕駛產業已全面轉型為無地圖端到端平台架構,無需依賴高清地圖,實現了低成本、高泛化能力和快速迭代的閉合迴路技術。同時,歐美主流汽車製造商仍採用傳統的依賴地圖架構,在迭代效率和場景適應性方面落後於中國。目前,小鵬汽車、理想汽車和鴻蒙智慧旅行聯盟(HIMA)的主要車型均標配無地圖NOA(網路化汽車)技術。根據華為統計,截至2026年4月20日,「千鈞ADS」累計行駛里程已正式突破100億公里。隨著車輛銷售的成長,累計行駛里程持續飆升。該演算法每週更新,在實施效率和成本控制方面均具有優勢。
中國是全球唯一擁有涵蓋「感測器—汽車晶片—域控制器—演算法—模擬標註—車輛整合」的完整自動駕駛產業鏈的國家。其技術實現週期短、產業化效率高,在全球擁有無可比擬的規模與成本優勢。核心硬體成本已大幅降低;LiDAR單價已從20萬元降至1000元,自動駕駛域控制器單價也已降至1000元。外國汽車製造商的先進自動駕駛策略高度依賴中國的供應鏈,這進一步強化了中國在全球自動駕駛產業中獨特的准入壁壘,其他國家則難以複製。
ResearchInChina 對 10 家汽車製造商和 7 家一級 ADAS 供應商的海外擴張和戰略計劃,以及歐洲、東協、拉丁美洲、中東、澳大利亞和紐西蘭等主要銷售市場 20 多款代表性車型的自動駕駛配置進行了詳細研究,並總結了四個關鍵見解。
洞察 1:隨著 L2+(不包括 LCA 和 NOA)成為海外自動駕駛部署的主流標準,硬體策略正在分岔成兩條路線:「主導大眾市場」和「為高階市場建立硬體基礎」。
對20餘款主流海外車型配置的分析表明,具備生命週期評估(LCA)的L2+系統是海外市場的核心智慧功能配置,佔比超過50%,並成為東協市場差異化的主要驅動力。基本版L2 ADAS作為一種成本績效實惠的基礎解決方案,而L1智慧駕駛目前僅在南美低階市場小規模推廣。整體而言,該技術正處於分階段發展階段:高階配置率先引入東協市場,L2系統在大眾市場佔主導地位,而最低限度的基礎配置則逐步引入南美市場。
就產品競爭而言,L1 功能價值較低,差異化不足,而海外競爭對手通常將 L1 作為車輛標配。在中國,車輛銷往全球,L2+(此處僅指加裝 LCA 的 L2 ADAS,不含 NOA)被廣泛採用為主流入門級標準。隨著海外基礎 ADAS 系統日趨同質化,LCA 以極低的邊際成本顯著提升用戶體驗,使其成為品牌優質化的關鍵,同時也為未來升級 NOA 奠定了硬體基礎。
在此背景下,中國汽車製造商在自動駕駛硬體的海外擴張方面採取了兩種截然不同的策略:
模式 1:先引入輕量級 ADAS,然後再增加高階功能(比亞迪、奇瑞、上汽 MG、躍躍欲試)
旨在擴大銷售的主流大眾品牌正採取「低成本市場定位和逐步升級」的務實策略。雖然包括LCA在內的L2/L2+自動駕駛功能已成為海外車型的標配,但NOA功能暫時不會包含在內。這些品牌優先考慮海外市場的銷售成長和用戶滲透,以精簡硬體並嚴格控制單車成本。一旦目標市場實施NOA法規、完成數據合規並成熟場景適配,將透過車型變更和OTA更新逐步解鎖高級自動駕駛功能。這最大限度地降低了海外認證風險和初期研發投入,並在銷售和盈利之間實現了有效平衡。
模式 2:硬體一次性安裝完畢,軟體功能透過 OTA 更新逐步解鎖(小鵬汽車)
小鵬汽車秉持前瞻性的高階策略,所有海外車型均採用毫不妥協的全硬體配置。標配國產圖靈晶片、NVIDIA Orin 顯示卡、800V 高壓平台以及聯發科 MT8676 處理器,確保感知功能擁有充足的運算能力和冗餘備份。車輛上市初期僅啟用 L2+ 功能。待完成 UN R171 認證、歐洲路測以及本地化調校後,將透過 OTA 方式推送 VLA 2.0 更新,立即解鎖高速公路和城市 NOA 功能,滿足所有場景需求。此舉避免了後續車款改款和重複認證帶來的成本。憑藉著世代硬體優勢,小鵬汽車在海外市場樹立了高階智慧品牌形象,擺脫了價格主導的過度競爭,並透過迭代軟體開發和附加價值服務建立了長期競爭優勢。
洞察 2:動力傳動系統架構正在多樣化,其中 EREV 和 HEV 脫穎而出,預計在 2026 年擴大海外銷售。
儘管純電動車(BEV)仍是海外擴張的主力軍,但預計中國汽車製造商將採用涵蓋純電動車(EV)、增程式電動車(EREV)、插電式混合動力車(PHEV)、混合動力車(HEV)和燃油車的矩陣式產品陣容,以有效應對全球多元化的市場環境。 2025年至2026年間,躍遷車在海外推出了純電動車和增程式電動車,長安汽車也宣布推出搭載藍芯引擎的全球混合動力車型,旨在利用混合動力車彌補海外產品的不足。
多元化動力傳動系統向海外擴張是市場需求、成本和政策三大因素共同作用的必然結果。從消費者角度來看,電池式電動車仍面臨充電基礎設施、低溫性能和續航里程有限的挑戰。而混合動力汽車則具備續航里程長、油耗低、適應性強等優勢,讓消費者無須擔憂。從商業角度來看,基於傳統底盤架構的混合模式邊際成本低,盈利。從政策角度來看,許多國家正在逐步放寬對內燃機(ICE)的禁令,為混合動力汽車進入海外市場創造了長期的市場機會。一旦混合動力汽車產品線完善,中國汽車製造商將能夠更好地應對全球各地不同的基礎設施、政策和消費市場。
洞見 3:聯合國 R171 S02 結束了海外自動駕駛部署的灰色地帶,2027 年將成為 NOA(無人駕駛自動技術)全球商業化的轉捩點。
2026年6月,聯合國歐洲經濟委員會(UNECE)WP.29大會通過了聯合國R171系列02號條例(關於駕駛員控制輔助系統(DCAS))的修正案,其中包括都市區車道變換、環島通行和都市區自動駕駛輔助系統(NOA)。全球首個關於都市區NOA的強制國際法規將於2026年底在歐盟生效。這項新法規將首次在國際規則層面承認都市區NOA的合法商業化地位,徹底消除長期困擾海外高級自動駕駛系統的「缺乏法律依據和功能受限」的灰色地帶。
此前,由於缺乏全球統一的高級自動駕駛法規,國內車型普遍未配備NOA功能,僅搭載基礎ADAS。隨著R171 S02的發布,業界將建立清晰的層級結構。換言之,基本L2 ADAS將成為暢銷低價車款的主流配置,而支援所有場景的NOA認證版本將搭載於高階豪華車款。 2027年將是全球高階自動駕駛大規模商業化的關鍵轉捩點。
全球汽車製造商都在競相抓住 2027 年引入 NOA(僅噪音接入)所帶來的機會。
在等待監管部門批准的同時,小鵬汽車正採取「技術優先」策略。該公司計劃於2026年第四季在歐洲啟動公共道路測試,並於2027年完成第二代VLA的全球交付。憑藉積極應對新法規的技術基礎,小鵬汽車已做好準備,在全球推出功能齊全的都市區NOA系統。
賓士計畫於2026年底正式進軍都市區駕駛輔助系統(NOA)市場,推出「MB.DRIVE ASSIST PRO」。都市區自動駕駛輔助系統最初將在德國主要城市推出,並計劃於2027年全面推廣。這將是一款基於「Orin」平台的先進都市區自動駕駛輔助系統,標誌著這家歐洲豪華品牌正式進軍先進的都市區自動駕駛輔助系統領域。
從 2025 年到 2026 年,特斯拉將繼續加速 FSD 的全球推廣,利用各領域的數據迭代來加強其海外自動駕駛基礎設施。
洞察 4:自動駕駛技術的海外擴張已從以產品貿易為中心的「1.0 時代」轉向以區域為基礎進行系統化的「2.0 時代」。
在1.0時代,採用的是單邊貿易模式。產品在國內完成,銷往海外,交貨後生命週期即告結束,缺乏持續迭代改進、在地化和合規性的封閉回路型。然而,都市區自動駕駛技術受場景、法規和數據等諸多因素的限制,標準化和即插即用難以實現。僅出口產品已不足以支撐高水準自動駕駛的實現。
該行業已在海外市場紮根,正式邁入「2.0時代」,從「硬體產品」擴展到「軟體+服務+解決方案」。其核心在於三大關鍵的在地化能力:本地化的研發機構、全數據流程的閉合迴路、本地化的技術運維以及自動駕駛基礎設施的建設。
其中,資料本地化是歐盟合規的最低要求。 GDPR和WP29嚴格禁止透過跨境資料回程傳輸訓練。同時,歐洲的道路權利、道路標誌和混合交通場景與中國之間存在顯著差異,僅依靠中國境內的遠端參數調整無法實現完全適應。因此,對於企業而言,在歐洲建立本地資料中心並建構「採集、標註、學習和迭代」的國內閉合迴路已不再是可選項,而是先進自動駕駛技術在海外合規商業化的必要前提。圍繞自動駕駛技術海外部署的競爭已從產品實力之爭演變為涵蓋海外研發現狀系統、資料合規框架和在地化迭代能力的系統性競爭。
案例研究:卓宇科技在德國自動駕駛市場的策略
自2025年起,卓宇全面實施「立足歐洲,服務歐洲」的本土化策略,並於2026年在德國布倫瑞克正式設立歐洲總部。公司建立了本地研發中心和區域測試資料中心,同時招募本地研發、車輛整合、合規和銷售人員,建構了完全本土化的營運組織架構,並使其基礎設施架構符合歐盟法規和OEM合作要求。
在業務營運方面採取兩種海外擴張策略。首先,順應中國自主品牌進軍全球市場的趨勢,為一汽紅旗、中國重汽等公司出口的車輛提供自動駕駛輔助系統(ADAS)服務。目前正與中國重汽聯合推進面向歐洲市場的商用車ADAS項目,雙方計劃同時在歐洲建立本地生產線。其次正積極進入歐洲整車廠(OEM)供應鏈,重點與福斯、BMW、賓士、奧迪等公司進行傳統OEM計畫合作。他們與歐洲大眾集團啟動了下一代智慧駕駛的概念驗證(POC)計畫。在該專案中選擇ID.Buzz作為檢驗車輛,並為其配備了Jimu System 2.0(感知系統)、智洲盲點補充雷達和Thor 11V1RSL感測器解決方案,已在歐洲完成了高速公路NOA、城市NOA和APA的實際道路測試。該專案目前處於概念驗證(POC)階段,他們計劃在 2026 年底前完成檢驗,然後進行大量生產和認證。
自動駕駛產品應用:卓宇科技已明確其在歐洲市場的商業化重點。高速公路NOA(網路自動駕駛)是量產和應用的首要任務,同時將傳統的L2 ADAS(進階駕駛輔助系統)作為基礎配置。到2026年,自動駕駛系統計畫演進為多模態統一表示平台模型(MFM)。這個自動駕駛系統利用單一平台模型實現現實世界中的統一感知和決策,原生兼容典型的歐洲場景,例如高速公路、城市街道、環島和路面電車。這不僅能夠使高速公路NOA在現有法規環境下實現量產和交付,也將為後續都市區NOA的商業化提供完整的技術基礎。該模型具有強大的跨區域泛化能力,單一架構即可無縫適配歐洲左舵駕駛市場以及英國和澳洲等右舵駕駛地區,實現全球範圍內的模型中心冗餘。
中國汽車製造商在自動駕駛領域的海外擴張方面實現了戰略飛躍,從被動的出口戰略轉向了積極主動的全球化戰略。國內股市的停滯迫使該產業尋求逐步向外部突破。成熟的無地圖基礎模型技術和完整的產業鏈正在創造全球性的、世代的技術優勢。聯合國法規R171 S02為全球自動駕駛高水準監管合規鋪平了道路。這三大因素共同作用,將使2026-2027年成為中國自動駕駛全球化轉型的重要時期。未來,產業競爭將不再僅圍繞車型或孤立的功能展開,而是圍繞著一個由全球產品矩陣、分層軟硬體策略、區域研發生態系統和合規資料系統所構成的綜合體系。中國汽車產業將正式從「產品出口時代」過渡到「智慧技術和行業標準出口時代」,重塑全球自動駕駛汽車的競爭格局。
Regulatory Breakthrough, Local System Establishment, OEMs Competing for NOA Layout: Overall Trends of China's Intelligent Driving Overseas Layout in 2026
Research on overseas intelligent driving layout: domestic stock-based competition has peaked, and globalization has become a must for the development of intelligent driving
The Chinese passenger car market has completely ended its incremental expansion cycle and entered the stage of stock-based competition. From 2020 to 2025, the compound annual growth rate of passenger car sales volume in China was only 2.34%, the share of fuel vehicles continued to shrink, the industry price war was normalized, and the involution further intensified. The increasingly saturated local market is no longer able to support OEMs' large-scale production capacity release and long-term high investment in intelligence at the same time. The industry's growth ceiling has emerged.
In sharp contrast to the sluggish growth of the local market, China's automobile exports have achieved explosive growth, evolving from a supplementary business in the past into a second core business pillar. The export volume soared from less than 1 million vehicles in 2020 to 7.098 million vehicles in 2025, a 7-fold increase in just six years; meanwhile, the export penetration rate jumped from 3.9% to 20.6%, and the export business had a compound annual growth rate of 48.14%. Overseas markets have become the core engine supporting the scale layout and growth of Chinese OEMs.
From a global industrial perspective, the next 3-5 years will be a critical window period for the reshaping of the global OEM landscape and the finalization of production capacity and brand echelons. Changan Automobile estimates that the global automotive market is expected to embrace 100 million vehicles in 2030, with the new energy vehicle penetration rate increasing to 45%. The Matthew Effect in the industry will continue to strengthen, the competition threshold will be higher, and distinctive echelons will be formed: for enterprises, the annual sales volume of 3-3.5 million vehicles represents the baseline for survival; the annual sales volume of 5-7 million vehicles indicates robust operational performance; the annual sales volume of 8-10 million vehicles allows them to build global competitive moats.
Reliance solely on the domestic market can no longer help enterprises survive. Global overseas layout has evolved from an "incremental option" into "the core growth engine for OEMs in the next phase of competition". The leading OEMs are comprehensively switching their development logic from "focusing on China and radiating overseas" to accelerating the iteration to a full-system global group with global R&D, manufacturing, channel and brand capabilities.
Many domestic OEMs have upgraded their strategic anchors: focusing on the domestic market + radiating overseas -> accelerating the transformation into global automobile groups with global R&D, manufacturing, channels and brand systems (2026-2030)
BYD aims to become the global leader in scale. It plans to sell over 5 million vehicles globally (more than 1.5 million vehicles overseas) in 2026, and 6 million vehicles in 2027 with a global market share of over 10%. In 2030, BYD will consolidate and enhance its leadership as the world's No.1 new energy vehicle seller by 2030.
Dongfeng Motor released the "Skyward Sailing" globalization plan in 2026, investing hundreds of billions of yuan in promoting the layout of the entire value chain overseas, following the three-step path of "products go global - production capacity goes global - industry goes global". It aims to sell 5 million vehicles worldwide in 2030, with new energy vehicles accounting for over 70% and overseas sales volume making up 40%.
Changan Automobile has upgraded its "Vast Ocean Plan 2.0" to promote a deep transition from "product trade" to "industrial roots" in terms of overseas business, with the goal of the strategic transformation from a vehicle exporter to a global automobile group. It strives to be among the top ten global OEMs in 2030, with the global sales volume exceeding 5 million vehicles. Its overseas sales target stands at 1.5 million vehicles, with an aspirational goal of 1.8 million vehicles, and it plans to build an overseas production capacity system including 800,000 vehicles.
Domestic intelligent driving is undergoing fierce competition, forcing technology to spill over globally
The domestic intelligent driving industry is enjoying rapid penetration in China, and the market has rapidly shifted from incremental penetration to stock-based competition. The intense involution has become one of the core driving forces for OEMs to deploy in broader overseas markets. Nowadays, intelligence has become a standard feature of passenger cars instead of an option.
Data shows that the market structure has seen a fundamental upheaval: the proportion of vehicle models with NL dropped from 49.6% in 2022 to 21.8% in Q1 2026; during the same period, the penetration rate of L2 and above autonomous vehicle models climbed from 34.7% to 69.8%. Basic L2 ADAS is a fundamental market, while high-end functions such as highway NOA and urban NOA have become the only high-growth segment starting from 2024.
Currently, competition in the domestic intelligent driving market across all price ranges has become saturated. The penetration rate of L2 and above intelligent driving functions has reached 80%-90% in high-end models priced above RMB200,000, and 60%-70% in mainstream models priced between RMB100,000 and RMB200,000. Meanwhile, advanced technologies continue to trickle down. The penetration rate of intelligent driving in entry-level models priced below RMB100,000 rose from 1.6% in 2022 to 16.7% in Q1 2026. Domestic intelligent driving has narrowed its growth space and compressed its profit margins. It has become an inevitable trend in the industry to export technology and capabilities overseas.
The core barriers are highlighted: China's intelligent driving industry has generational advantages in both technology and industry chain, and boasts global output capabilities
Compared with the European, American, Japanese and South Korean markets, China's intelligent driving industry has built a three-fold generational advantage spanning technical architecture, engineering and complete supply chain. It stands as one of the few core domestic industries capable of achieving technical export with higher-dimensional superiority to overseas markets.
China is significantly ahead of the world in terms of the popularization and iteration of high-level intelligent driving. China's intelligent driving industry has fully switched to the mapfree end-to-end foundation model architecture, without depending on HD maps, achieving a technology closed loop featuring low cost, strong generalization and fast iteration. However, mainstream OEMs in Europe and the United States still use the traditional map-dependent architecture, and lag behind in iteration efficiency and scenario adaptability. Currently, major vehicle models such as those of XPeng, Li Auto, and Harmony Intelligent Mobility Alliance (HIMA) are equipped with mapfree NOA as standard. Huawei said that as of April 20, 2026, the total mileage involved with Qiankun ADS had officially exceeded 10 billion kilometers. As the sales volume of cars increases, the total mileage continues to soar. The algorithm is iterated on a weekly basis, forming dual advantages in implementation efficiency and cost control.
China is the only country in the world that has a complete intelligent driving industry chain of "sensors - automotive chips - domain controllers - algorithms - simulation & annotation - vehicle integration". It has a short technology implementation cycle, high industrialization efficiency, and unique scale and cost advantages in the world. Core hardware has seen significant cost reduction: the unit price of LiDAR has plummeted from RMB200,000 to RMB1,000, and the unit price of intelligent driving domain controllers has been also down to RMB1,000. Overseas OEMs' high-level intelligent driving layout is highly dependent on China's supply chain, further consolidating China's global intelligent driving industry barriers that cannot be replicated.
ResearchInChina summarizes 4 major insights through in-depth research on the overseas layout and strategic planning of 10 OEMs and 7 Tier 1 ADAS suppliers, as well as intelligent driving configurations of over 20 typical vehicle models in major sales markets such as Europe, ASEAN, Latin America, the Middle East, Australia and New Zealand.
Insight 1: L2+ (LCA, excluding NOA) becomes the mainstream standard for overseas intelligent driving layout, with hardware strategies diverging into two routes: "dominating mass market" and "laying hardware foundation for high-end market".
Based on the configurations of 20+ mainstream overseas vehicle models: L2+ systems with LCA have become the core intelligent configuration for overseas markets, accounting for more than 50%, as the main force for differentiated competition in the ASEAN market; basic L2 ADAS serves as a basic cost-effective solution; L1 intelligent driving is only deployed on a small scale in the South American low-end market. A tiered strategy is adopted overall: high-end configurations for ASEAN, L2 dominating the mass market, and minimal basic specifications for South America.
As for product competition, L1 has low functional value and insufficient differentiation, while overseas competitors have generally included L1 as a standard feature. Chinese vehicle models going global use L2+ (here, L2+ vehicle models refer only to those adding LCA to L2 ADAS without NOA involved) as the mainstream entry-level standard. Against the backdrop of increasingly homogenized basic ADAS overseas, LCA has achieved a leap in experience at extremely low marginal cost, becoming crucial for brand premium, while simultaneously laying the hardware foundation for future NOA upgrades.
Against this background, Chinese OEMs have formed two completely different overseas layout strategies for intelligent driving hardware:
Mode 1: Lightweight ADAS first, high-end functions later (BYD, Chery, SAIC MG, Leapmotor)
Mainstream popular brands pursuing high sales volume adopt a pragmatic strategy of "low-cost market positioning and progressive upgrades". Overseas vehicle models come standard with L2/L2+ including LCA, but does not carry NOA for the time being. In order to streamline the hardware and strictly control the cost of a single vehicle, priority is given to overseas market distribution and user penetration. After NOA regulations are implemented in the target market, data compliance is completed, and scenario adaptation is mature, high-level intelligent driving capabilities will be gradually unlocked via iterative vehicle models or OTA updates, minimizing overseas certification risks and early R&D investment pressure, and achieving an effective balance between sales volume and profitability.
Mode 2: Hardware is pre-installed in one step, and software is unlocked in batches via OTA updates (XPeng)
XPeng adopts a forward-looking high-end layout strategy. All its overseas vehicle models come with full hardware specifications without compromise. They are standardly equipped with Turing chips made in China, NVIDIA Orin, 800V high-voltage platforms and MediaTek MT8676, reserving ample computing power and perception redundancy. Only L2+ functions will be enabled at the initial launch of the vehicle models. After completing UN R171 certification, European road tests and localized model tuning, VLA 2.0 will be rolled out via OTA to unlock all-scenario highway and urban NOA in one go. This mode avoids costs incurred by subsequent vehicle model revisions and repeated certifications. Leveraging generational hardware advantages, it secures a premium intelligent brand positioning overseas, breaks away from price-driven cutthroat competition, and builds long-term moats for software iteration and value-added services.
Insight 2: Powertrain architectures are diversified, and EREVs/HEVs stand out as breakthroughs for overseas sales growth in 2026
BEVs are still the main force going overseas, but Chinese OEMs will adopt a full-domain matrix layout of EVs+EREVs+PHEVs+HEVs+fuel vehicles to accurately adapt to the global differentiated market environment. From 2025 to 2026, Leapmotor launched BEVs and EREVs overseas, and Changan unveiled a global HEV model fitted with the BlueCore engine to make up for the shortcomings of overseas products with HEVs.
The overseas rollout of diversified powertrains is an inevitable result driven by three factors: market demand, cost and policies. From the consumer side, battery electric vehicles remain constrained by charging infrastructure, low-temperature performance and long-distance range limitations. Hybrid vehicles offer range anxiety-free driving, low fuel consumption and adaptability to all usage scenarios. From the corporate side, hybrid models built on conventional chassis architectures come with lower marginal costs and better profitability. From the policy side, many countries have relaxed timelines for ICE bans, creating a long-term market window for hybrid vehicles to enter overseas markets. After the hybrid matrix is completed, Chinese OEMs can cover different infrastructure, different policies, and different consumer markets around the world.
Insight 3: UN R171 S02 ends the gray zone of overseas intelligent driving layout, and 2027 marks the turning point for global NOA commercialization
In June 2026, the UNECE WP.29 plenary session adopted updates including UN R171 Series 02 for Driver Control Assistance Systems (DCAS), embracing automatic lane changes in urban areas, roundabout passage, and urban NOA. The world's first mandatory international regulations for urban NOA will be mandatory in the EU at the end of 2026. The new regulations recognize the legal commercial status of urban NOA at the level of international rules for the first time, completely ending the gray zone of "no legal basis and locked functions" for high-level intelligent driving systems overseas for many years.
Previously, due to the lack of global high-level intelligent driving regulations, domestic vehicle models generally omitted NOA functions but only retained basic ADAS. With the launch of R171 S02, the industry will form a clear layered approach: the basic L2 ADAS will prevail in best-selling low-priced vehicles, and the certified version of all-scenario NOA will be seen in high-end premium vehicles. 2027 will mark a key turning point for the large-scale commercialization of global high-level intelligent driving.
Global OEMs are simultaneously racing to seize the NOA window period in 2027
XPENG adopts a technology-first strategy to await regulatory clearance. It will launch European road tests in Q4 2026 and complete the global delivery of the second-generation VLA in 2027. Supported by a technology foundation pre-adapted to the new regulations, XPENG will achieve the overseas rollout of fully-featured urban NOA.
Mercedes-Benz plans to launch MB.DRIVE ASSIST PRO at the end of 2026 and officially enter the urban NOA arena. It will initially launch urban NOA in core cities in Germany and roll it out comprehensively in 2027. It will realize complex urban NOA based on Orin, marking the European luxury brand's entry to high-level urban NOA.
Tesla continues to accelerate the global rollout of FSD in 2025-2026, relying on full-domain data iteration to consolidate its overseas intelligent driving base.
Insight 4: Overseas layout of intelligent driving has moved from the 1.0 era of product trading to the 2.0 era of territorial systematization
The one-way trade model was adopted in the 1.0 era: products were finalized domestically, sold overseas, and the lifecycle terminates upon delivery, with no ongoing iteration, localized adaptation or compliance closed-loop. However, urban NOA is a technology heavily bound by scenarios, regulations and data. It cannot be standardized and plug-and-play. Simply exporting products can no longer support the implementation of high-level intelligent driving.
The industry has officially entered the 2.0 era of taking root in overseas markets, extending from "hardware products" to "software + services + solutions". The core relies on three major local capabilities: local R&D organization, data full-link closed loop, localized technical operation and maintenance, and the construction of intelligent driving infrastructure.
Wherein, data localization is the rigid bottom line of EU compliance: the GDPR and WP29 strictly prohibit cross-border data backhaul training. At the same time, European road rights, road signs, and traffic mixed scenarios are hugely different from Chinese ones, making it impossible to rely on China's domestic remote parameter adjustment to complete adaptation. Therefore, building a local data center in Europe and building a domestic closed loop of "collection-labeling-training-iteration" is no longer an optional advantage for enterprises, but a necessary prerequisite for the overseas compliance commercialization of high-level intelligent driving. Competition in overseas intelligent driving layout has evolved from rivalry over product strength into systematic competition encompassing overseas R&D ecosystems, data compliance frameworks and localized iteration capabilities.
Case study: Zhuoyu Technology's strategy in the German intelligent driving market
Since 2025, Zhuoyu has fully implemented its deep localization strategy of "In Europe, for Europe," and officially opened its European headquarters in Braunschweig, Germany in 2026. It has built its own local R&D center and regional testing data center, simultaneously recruiting local talents for R&D, vehicle integration, compliance, and business positions, establishing a complete localized operating organization, and adapting its underlying architecture to EU regulations and OEM cooperation requirements.
On the business side, it adopts a two-pronged approach in overseas layout: First, following the wave of Chinese independent brands going global, it provides intelligent driving support for vehicles exported by FAW Hongqi, Sinotruk, and others; currently, it is advancing a commercial vehicle ADAS project for the European market with Sinotruk, and both parties are simultaneously planning the establishment of a localized production line in Europe. Second, it enters the European OEM supply chain, focusing on traditional OEM projects such as Volkswagen, BMW, Mercedes-Benz and Audi, and has already launched the next-generation intelligent driving POC (Proof of Concept) with the Volkswagen Group in Europe. The project selects ID.Buzz as the verification vehicle, equipped with Jimu System 2.0 (perception system), Zhizhou Blind-spot Filling Radar, and Thor. With an 11V1RSL sensor solution, it has completed real-vehicle road tests on highway NOA, urban NOA, and APA in Europe. The project is currently in the POC stage and plans to complete verification by the end of 2026, followed by mass production and designation.
Intelligent driving product application: Zhuoyu Technology has clearly defined commercialization priorities in the European market: highway NOA will be prioritized for mass production and application, while conventional L2 ADAS will be deployed simultaneously as a basic configuration. In 2026, the intelligent driving system will iterate to a multi-modal unified representation foundation model (MFM). Relying on a single foundation model to complete unified perception and decision-making of the physical world, the intelligent driving system is natively compatible with typical European scenarios such as highways, urban roads, roundabouts, trams, etc. It can not only directly match the current regulatory environment to conduct highway NOA mass production and delivery, but also provide a complete technical foundation for subsequent urban NOA commercialization. The model has strong cross-zone generalization capabilities, and can seamlessly adapt to the left-hand drive markets in Europe and right-hand drive regions such as the United Kingdom and Australia in one architecture, achieving model subject multiplexing on a global scale.
Chinese OEMs have completed a strategic leap in the overseas layout of intelligent driving, shifting from a passive export strategy to active global strategy. The involution of the domestic stock-based market has forced the industry to seek incremental breakthroughs externally. The mature mapfree foundation model technology and the complete industry chain have created a global generational technological advantage. The UN R171 S02 has opened a global compliance window for high-level intelligent driving. With the combined benefits of these three factors, 2026-2027 will be the transformative era for the globalization of China's intelligent driving. In the future, industry competition will no longer only focus on vehicle models or isolated functional capabilities, but center on a comprehensive system consisting of global product matrix, layered hardware-software strategies, localized R&D ecosystems, and compliant data systems. China's automotive industry will officially move from the "product export era" to the "intelligent technology and industry standard output era", reshaping the global autonomous vehicle competition landscape.