Abstract: To address the problems of complex wiring and installation, low spatial coverage of single-point monitoring, and limited accuracy under non-uniform deformation in conventional roadway surrounding rock deformation monitoring, a technical framework is established for high-precision quantitative analysis and differentiated synergistic control of mining-induced roadway deformation over the full cross-section and entire lifecycle. A fine-grained roadway deformation analysis method based on spatiotemporal 4D point clouds is proposed. A portable mobile 3D laser scanning system integrating simultaneous localization and mapping (SLAM) and an inertial measurement unit (IMU) is developed to acquire multi-temporal high-fidelity 3D point clouds of the roadway. Based on this, a Two_Stage hierarchical deformation optimization algorithm is designed. In the first stage, normal vectors and cylindrical projection are used for coarse deformation computation to rapidly locate key deformation regions. In the second stage, strict geometric constraints and displacement vectors are employed to perform refined computation for core points exceeding deformation thresholds, thereby mitigating the geometric-mean weakening effect and improving the analysis accuracy for non-uniformly deformed surfaces. Finally, by incorporating distributed anchor bolt (cable) stress monitoring data, a support-rock synergistic bearing efficiency index is constructed to enable quantitative evaluation of support effectiveness. Under different deformation gradient experient, an average absolute error of 0.0483 m is achieved by the Two_Stage algorithm. Compared with the multiscale model-to-model cloud comparison (M3C2) and cloud-to-cloud (C2C) methods, the measurement accuracy is improved by 68.7% and 82.6%, respectively. It is shown by field applications that the deformation of small coal pillar roadways exhibits pronounced spatiotemporal anisotropy: spatially, an asymmetric pattern dominated by floor heave is presented, with more severe deformation on the small coal pillar side; temporally, two stages are characterized by intense advance influence within about 150 m ahead of the working face and post-mining lagging deformation. It is indicated by quantitative analysis that the SEI value on the small coal pillar side is about 0.25 MPa/mm, only one-third of that on the solid coal side, and a mechanical decoupling phenomenon of “high deformation and low stress” on the small coal pillar side is revealed. The large asymmetric deformation on the small coal pillar side is mainly attributed to reduced anchorage system bearing efficiency caused by surrounding rock fragmentation. Based on the coupled deformation-stress monitoring results, an asymmetric synergistic control technology is developed, combining roof cutting for pressure relief to optimize the stress environment, floor grouting to suppress heave, and deep reinforcement with long anchor cables. It is demonstrated by field monitoring that lateral stress concentration toward the roadway is suppressed and the synergistic bearing relationship between shallow and deep rock masses is reshaped by this scheme, thereby forming a closed-loop technical route from full-section deformation monitoring and instability diagnosis to targeted intelligent control.