"根植升级"是本土代工企业升级前沿的新动态。突破主流升级观的"极化"局限，基于"根植性"假设，整合自我中心网络治理与企业能力理论，运用长三角地区本土代工企业在国内、欧洲、美国、巴西、俄罗斯等海内外根植升级案例与样本数据，采用科学研究范式与多种实证方法，研究根植升级机理：（1）根植升级跨案例研究，探索根植网络治理、升级能力与租金攫取等概念的结构"黑箱"与因果链；（2）构建四个根植升级机理模型并大样本检验：根植网络治理模型、根植学习模型、根植网络跨区多向能力溢出模型、"根植-嵌入"杠杆效应模型；（3）基于共同演化范式的根植升级路径与动态演化机理跨案例跟踪研究；（4）根植升级策略、政策与系列应用工具开发。本土代工企业"根植升级"的提出具有主流突破性，根植升级机理理论具有本土原创性，研究结论对本土代工企业寻求升级具有直接应用价值，部分根植升级成效显著的本土案例，必将成为升级标杆和政府决策的直接依据

Upgrading by Rooting is the latest trend of local OEM/ODM firms upgrading. Breaking through the polarized shortage of existing upgrading view and adopting rootingness hypothesis, this study combines ego network governance and firm's capability theories to research the mechanism of upgrading by rooting, based on the research paradigm of local original theory construction and various experimental research methods, with cases and samples of local OEM/ODM firms in Yangtze River Delta upgrading successfully by rooting abroad in Europe,USA,Brasil,Rusia or domestically :(1) Cross-cases study on upgrading by rooting. Explore the concept structure of rooting network governance, upgrading competency and rent appropriation performance, and their relationships;（2）Construct four mechanism models of upgrading by rooting: Rooting network governance model, Learning by rooting model, Model of multi directional capabilities spillover across various blocks of the rooting network, and Leverage effect model between rooting and embedding.(3)Longitudinal cross-cases study on the trajectories and evolutional dynamics of upgrading by rooting mechanism based on the paradigm of co-evolution analysis.（4）Develop strategies, policies and analytic tools for upgrading by rooting. The concept of upgrading by rooting is a sharp innovation to the traditional upgrading theory, and the theory of the mechanism is an original research in the field of upgrading. The conclusions will be dramatically valuable to local OEM/ODM firms' upgrading, some of the cases with outstanding upgrading performance will be benchmarks in expectation and be beneficial to policies decision.