The field of soft robotics is rapidly growing as new and exciting technologies seek to improve upon and replace rigid robotic systems. Soft robotic systems provide a unique opportunity to embed actuation and sensing directly into the materials of a soft robot, enabling distributed intelligence, computation, and communication capabilities along the robot body. One area of interest where soft devices excel is in surgical robotics. Rigid robotic systems used in surgical robotics lack the dexterity, compliance, and sensor feedback for optimized surgical procedures. Soft robots represent a promising technology in this field because they are constructed from compliant materials and can thus safely interact with delicate anatomical structures, preserving their physiological function. In surgical applications, a soft robot with proprioceptive capabilities to improve closed-loop feedback control and exteroceptive capabilities to monitor contact forces within a patient provides an opportunity to enhance navigational capabilities and allow for greater dexterity of devices. Ultimately, these advancements push toward improved patient safety and recovery times as well as making procedures easier for surgeons. The investigation into these technologies has been embodied in projects seeking to improve upon minimally invasive surgical procedures (MIS) such as bronchoscopy, laparoscopy, and colonoscopy. First, the design, modeling, fabrication, and testing of a soft robotic bronchoscope with enhanced vision sensing and integrated diagnostic and interventional capabilities is presented. Lung cancer is one of the deadliest forms of cancers and is often diagnosed by performing biopsies with the use of a bronchoscope. However, this diagnostic procedure is limited in its ability to explore deep into the periphery of the lung where cancer can remain undetected. The small diameter (2.4~mm) of the device allows navigation in branches deeper in the lung, where current devices have limited reachability. We performed mechanical characterizations of the robotic platform, including blocked force, maximum bending angle, maximum angular velocity, and workspace, and assessed its performance in in vitro and ex vivo experiments. We developed a computer vision algorithm, and validated it in in vitro conditions, to autonomously align the robot to a selected branch of the lung and aid the clinician (by means of a graphic user interface [GUI]) during navigation tasks, and to perform robotic-assisted stabilization in front of a lesion, with automated tracking and alignment. Then, we developed a roughness tuning strategy for the fabrication of soft optical waveguide sensors (WG) to achieve shape sensing and contact recognition within a single multi-modal sensor. We integrated the sensors into a fully soft continuum robot consisting of a multi-directional bending module and a gripper. The robot module integrates two WGs for 3D shape sensing and three soft pneumatic actuators to steer the robot in all directions. The gripper embeds two soft pneumatic actuators to deploy itself, two soft pneumatic actuators to control grasping, and two WGs, one in each jaw of the gripper, with tuned roughness to monitor both gripper tip positions and subsequent occurrence of contact with an object. The accuracy of our sensing strategy is shown in validation experiments and an in vitro experiment is conducted in a mock laparoscopic environment to exhibit our robot’s functionality in a surgical scenario. Our robot can perform a peg transfer test in an in vitro laparoscopic environment while monitoring the shape of the robot and the force exerted on objects of varying sizes and outputting the information to a surgeon via a GUI. Our soft continuum robot can be utilized in a variety of pick-and-place applications where the integration of soft optical sensors can improve the accuracy of the control of soft robotic arms. Further, we determined that proper tuning of the sensors can result in increased sensitivity across a large range of curvatures and enable a multi-modal sensor response, which allows for a reduced number of total sensors in the robot. Lastly, a soft robotic sleeve to provide sensor feedback and additional actuation capabilities to improve safety during navigation in colonoscopy is presented. Colonoscopy is the gold standard for colorectal cancer diagnosis; however, limited instrument dexterity and no sensor feedback can hamper procedure safety and patients’ acceptance. The robot can be mounted around current endoscopic instrumentation as a disposable ``add-on'', avoiding the necessity for dedicated or customized instruments and without disrupting current surgical workflow. We focus on design, finite element analysis modeling, fabrication, and experimental characterization and validation of the soft robotic sleeve. The device integrates soft optical sensors to monitor contact interaction forces between the colon and the colonoscope as well as soft robotic actuators that can be automatically deployed if excessive force concentration is detected, to guarantee pressure redistribution on a larger contact area of the colon. The system can be operated by a surgeon via a GUI that displays contact force values and enables independent pressurization of the soft actuators upon demand, in case deemed necessary to aid during navigation or distend colon tissue. These novel soft sensing systems have succeeded in providing enhanced capabilities to soft robots in MIS and simultaneously addressing key challenges in robotics, such as navigating in these complex and unstructured environments within the body and providing effective force outputs at the distal tip of soft systems. Further, the advancements in the tunability of soft optical sensors paves the way toward miniaturized devices and better closed-loop control of soft robots. Providing contact force recognition to continuum robots also pushes toward improving the capabilities of haptic feedback systems that can be integrated with soft sensors and soft continuum robots. Lastly, as described, the low cost and scalable nature of the system presents opportunities for incorporation into a variety of MIS procedures where enhanced dexterity and sensor feedback is valuable for surgeons.