At present, tunnel boring machine (TBM) tunneling depends heavily on manual operation, resulting in limited automation and intelligence. To further improve tunneling safety, quality, and efficiency, an automatic cruise control system for TBM tunneling based on big data and artificial intelligence technologies are designed and developed. The system employs multisource sensor fusion to perceive surrounding rock conditions in real time, enabling autonomous adjustment of tunneling parameters such as cutterhead rotation speed and thrust velocity. Meanwhile, by integrating real-time TBM position and attitude data provided by a high-precision guidance system, a deep transfer learning neural network is used to achieve real-time tracking of the designed alignment and dynamic adjustment of the TBM attitude, thereby enabling automatic deviation correction and directional control. In addition, the system realizes fully automated collaborative control of the stepping process. The system has been successfully applied in the Beishan No. 1 TBM project of the Beishan underground laboratory. Application results demonstrate that (1) under unmanned conditions, the TBM autonomously identifies surrounding rock conditions and dynamically matches tunneling parameters, successfully achieving continuous downhill tunneling under a 10% longitudinal slope and completing a horizontal turn with a radius of 255 m; and (2) the system achieves highly autonomous operation throughout core processes, including excavation, deviation correction, and stepping, ultimately controlling the deviation of the entire tunnel axis within ±50 mm and completing each single-cycle stepping process within no more than 5 min. The successful development and application of this system confirm the feasibility of long-term autonomous cruising of TBMs under complex construction conditions and provide a practical engineering example for advancing intelligent tunneling technology.