I. Introduction
Research on robots with rigid components has a long history. Designing and controlling soft robots for a variety of tasks - from crawling on surfaces to carrying different payloads - is just beginning to be understood [1], [2]. Soft robots are attractive because: (1) they are deformable and can closely mimic animal behavior; (2) they are soft and thus more collision-resilient than traditional rigid-body robots; (3) they can perform tasks that rigid robots cannot perform [3]–[10]. The actuators in a soft robot can be activated electrically, thermally, or pneumatically [11]. Many factors affect the locomotion of soft robots, including the actuation patterns of the actuators, the weight distribution across the soft robot body, and the capability of untethered operation. Some of the above demonstrations are tethered hence only consider the impact of weight distribution in the context of tethered operation. Some other designs have limited control degrees of freedom (DOF). Research on the energy efficiency, and actuation-locomotion co-design is still at the early stage [12].