<p>Legged robots have considerable potential for traversing unstructured situations; nonetheless, their inflexible frameworks often constrain adaptability and obstacle negotiation. The study article presents a revolutionary Soft Tri-Legged Robot (STLR) that improves movement and obstacle-avoidance skills by using a bio-inspired pneumatic artificial muscle (Bubble Artificial Muscles) and a bio-inspired tactile sensor (TacTip). The STLR is activated by BAMs, which are flexible, pneumatic-driven actuators that provide fine control over forward, backward, and steering movements. Obstacle identification and avoidance are facilitated by the TacTip sensor, which delivers tactile input for traversing unstructured terrains. We delineate the mechanical features of the BAMs, assess the functionality of the robot’s legs, and elaborate on the incorporation of the tactile sensing system. Experimental results demonstrate that the STLR can effectively achieve multi-directional flexible movement and obstacle avoidance through a cross-modal perception-actuation mechanism. This study highlights the promise of soft robotics for search and rescue, medical aid, and autonomous exploration, while delineating difficulties and opportunities for future improvements in functionality and efficiency.</p>