達摩院王剛首次對外解讀阿里自動駕駛佈局，阿里無人車配送快遞已突破100萬單

{"type":"doc","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"blockquote","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"自動駕駛技術的發展不能再靠講故事、畫大餅和資本燒錢的方式來推進，業界需要看到清晰的商業模式"}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":3},"content":[{"type":"text","text":"阿里自動駕駛實現規模化落地，無人車配送快遞突破100萬單"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"9月27日，阿里巴巴集團副總裁、達摩院自動駕駛實驗室負責人王剛宣佈，達摩院自研的L4級自動駕駛產品、末端物流無人車“小蠻驢”已落地全國22個省份，累計配送訂單超100萬單，爲20多萬人送過快遞。3年後，小蠻驢車隊規模將達1萬輛，有望日均配送包裹100萬件。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https:\/\/static001.infoq.cn\/resource\/image\/f2\/ae\/f269ecf82870cb85367f6877caea4eae.png","alt":null,"title":"","style":[{"key":"width","value":"75%"},{"key":"bordertype","value":"none"}],"href":"","fromPaste":false,"pastePass":false}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"小蠻驢是阿里爲“最後三公里”研發的自動駕駛產品，服務於快遞、外賣、生鮮及各類即時配送需求。2020年9月問世後，達摩院與菜鳥合作，率先針對快遞配送形成服務閉環，將高校、社區、園區的快遞按用戶預約時間送貨到樓。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https:\/\/static001.infoq.cn\/resource\/image\/2d\/5f\/2d35969ba983fea2fa7fa9f9b262ce5f.png","alt":null,"title":"","style":[{"key":"width","value":"75%"},{"key":"bordertype","value":"none"}],"href":"","fromPaste":false,"pastePass":false}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":"center","origin":null},"content":[{"type":"text","marks":[{"type":"italic"}],"text":"“小蠻驢”"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"這是王剛首次對外解析阿里巴巴自動駕駛戰略佈局。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"2017年達摩院成立，自動駕駛實驗室承擔起阿里集團自動駕駛技術的整合統一任務，由王剛帶隊組建研發力量，開始面向應用的技術探索。王剛強調，阿里將持續長期投入自動駕駛。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"自動駕駛技術發展了三十多年，終於開始大規模應用，進入普通人的日常生活。王剛說：“自動駕駛已經進入產品驅動的發展階段。小蠻驢的100萬單只是一個起點，它爲自動駕駛技術的進一步發展和落地打下了很好的基礎。”"}]},{"type":"heading","attrs":{"align":null,"level":3},"content":[{"type":"text","text":"100萬訂單背後"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"在接受InfoQ等媒體採訪時，王剛表示，小蠻驢快遞訂單突破100萬意味着自動駕駛的規模化應用實現了從0到1的突破，在產能爬坡最艱難的早期階段，達摩院自動駕駛實驗室逐步克服了技術、成本、供應鏈等多方面的挑戰，驗證末端物流這一垂直賽道。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"bulletedlist","content":[{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"技術挑戰"}]}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"無人車技術涵蓋算法、算力、硬件、系統等多個維度，算法能力最能體現無人車的智能水平。算法研發依賴數據驅動，需要處理大規模的場景數據、運營數據。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"達摩院自動駕駛實驗室通過強化學習、智能仿真等手段，掌握了一套對數據進行結構化分析和處理的能力，比如，提出了“小前臺、大中臺”理念，首創自動駕駛“算法中臺”AutoDrive，讓計算機自動學習並找到適合每種算法模型的結構和參數，用更符合計算機思維規律而非人類思維規律的方式處理數據，“以計算換智能”。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"bulletedlist","content":[{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"場景挑戰"}]}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"末端物流場景遠比人們想象的複雜。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"不同於規則明確、參與者有限的公開道路“結構化場景”，末端環境是典型的“非結構化場景”，交通參與者包括人、小動物、非機動車、機動車、各種路障和臨時障礙物等，沒有規範的車道設計和通行規則，無人車與任何道路參與者的交互都是複雜的博弈問題，比如窄路會車，無人車需要辨別對方意圖並作出“你進我退”、“你退我進”等的行車決策。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"bulletedlist","content":[{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"成本挑戰"}]}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"以大規模應用爲目標的自動駕駛產品，如果不能將成本降低到量產水平，一切都是紙上談兵。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"從末端場景的社會需求來說，“成本可控”也是無人車的應用前提之一。我們通過核心算法自研+核心硬件深度定製的方式，大幅降低了無人車的研發製造成本。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"小蠻驢的核心傳感器、底盤和嵌入式計算平臺全部採用深度定製，達摩院高性能、低功耗、低成本的嵌入式異構計算單元，能以1\/3算力的達到同等智能水平。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"此外，達摩院自動駕駛實驗室還逐步克服了無人車量產和規模化運營過程中遇到的供應鏈、運維等多種挑戰，在從0到1的階段，逐步推動產能爬坡。"}]},{"type":"heading","attrs":{"align":null,"level":3},"content":[{"type":"text","text":"直接定位L4級路線，“直奔”商業驗證"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"自動駕駛技術百家爭鳴，總體來說分爲兩大路線：路線一：從L2到L4的縱向技術迭代路徑，比如特斯拉。從輔助駕駛技術起步，固定場景，向上逐步實現技術迭代。路線二：從垂直場景L4到泛化場景L4，橫向場景拓展路徑。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"不同於從L2到L4的縱向技術迭代路線，阿里自動駕駛走的是L4級橫向場景拓展路線，自始至終定位L4級技術，從垂直領域起步，在非載人駕駛場景下完成技術累積，逐步切換到無人卡車、乘用車、移動機器人等泛化場景。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"“自動駕駛技術的發展不能再靠講故事、畫大餅和資本燒錢的方式來推進，業界需要看到清晰的商業模式”，王剛在接受媒體採訪時表示，“因此，阿里選擇了一條“技術高打”、“場景下沉”的路線，希望少走彎路，且快速完成末端物流賽道的商業驗證”。"}]},{"type":"heading","attrs":{"align":null,"level":3},"content":[{"type":"text","text":"阿里無人卡車“大蠻驢”已啓動研發，與小蠻驢共享同一技術框架"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"王剛表示，面向公開道路的無人卡車“大蠻驢”也已啓動研發。“大蠻驢”是一款適用於城配物流的輕型卡車，應用場景包括城市快遞網點-快遞末端節點之間的物流配送等。目前，達摩院和菜鳥合作，已展開無人輕卡路測。阿里計劃在3年內讓無人卡車出現在上萬條城市配送路線上，開展常態化運營。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"比起物流車，無人卡車的研發面臨着更大的技術挑戰，但小蠻驢已經沉澱的技術能力和場景數據爲無人輕卡的快速研發奠定了基礎。大小蠻驢共享同一技術框架，達摩院自研的自動駕駛機器學習平臺AutoDrive、雲上智能仿真測試平臺、嵌入式計算單元以及深度定製傳感器技術，都已順滑“平移”到了無人卡車的技術棧。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https:\/\/static001.infoq.cn\/resource\/image\/bb\/76\/bbdb59786be2449df0f7b9f1be891576.png","alt":null,"title":"","style":[{"key":"width","value":"75%"},{"key":"bordertype","value":"none"}],"href":"","fromPaste":false,"pastePass":false}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":"center","origin":null},"content":[{"type":"text","marks":[{"type":"italic"}],"text":"小蠻驢一年成績"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"基於小蠻驢的技術底盤，達摩院正在開發多種自動駕駛產品，如具備自主移動能力的電力巡檢機器人。首批量產的巡檢機器人，以服務機器人的成本實現了工業機器人的性能，即將批量進入電力系統，取代人工開展電力巡檢工作。"}]}]}