Existing tunnel boring machine (TBM) construction presents certain shortcomings. These include difficulty in comprehensive perception of information, poor timelines of information transmission and storage systems, significant effects of traditional data processing methods on the timeless of intelligent decision-making, and poor applicability of decision-making models and control strategies. In addition, the integration level of perception, decision-making, and control should be further improved. Therefore, a cross-platform deployable intelligent tunnelling robot system with closed-loop intelligent control functions of a “comprehensive perception, dual-driven decision-making, and composite intelligent control” is developed. Based on fieldbus, communication, database, cloud computing, and advanced exploration technologies, a multi-source information perception and integrated management platform based on a two-layer architecture is built to achieve the comprehensive perception of tunnelling information. In addition, an optimal decision-making method of the particle swarm optimisation (PSO) algorithm is simultaneously proposed for the minimum decision-making of tunnelling specific energy for scientific analyses and decision-making. A composite intelligent control strategy comprising multimodal and expert experienced learning control strategies is designed to achieve the control of conventional and unfavourable geological sections, respectively. Engineering cases verified the effectiveness and reliability of the intelligent tunnelling robot system. The research results not only provide new ideas and technical means for achieving the less-manned, unmanned, and intelligent tunnelling construction of deep-buried long tunnels but can also be promoted owing to its universality.