Annelid-Inspired Locomotion using Bi-Stable Origami Towers

Nature’s Inspirations

The earthworm and many species of snakes use rectilinear locomotion to propel themselves forward using a combination of radial and longitudinal abdominal muscles. The radial muscles contract to lift a segment of the body while longitudinal abdominal muscles expand to propel the anterior body segments forward. The radial muscles then expand to anchor the segment to the ground while the abdominal muscles contract to propel the posterior segments forward. This type of locomotion is very compact and ideal for traversing constricted spaces. The large amount of surface area in contact with the ground in conjunction with the segmented nature of the motion make rectilinear locomotion also optimal for traversing rough and complex terrain.

Engineering Challenges

Most mobile robots lack the ability to traverse complex terrains and compact spaces due to the inability to adapt to their surroundings. Traditionally, robots have consisted largely of mechanical hardware, making them large, bulky, and generally unsuitable for harsh and unpredictable environments. While the development of soft robotics has provided means for increased maneuverability and adaptability, it has brought new challenges such as decreased durability and lack of autonomy due to modeling complexity. For a mobile robot to be useful for applications such as search and rescue, intelligence, and surveillance missions, it must be autonomous and able to withstand harsh and complex terrains. The ability of a robot to autonomously segment in the field would allow missions to continue even if part of the robot were stuck or disabled and prevented from continuing.

BAM Approach

In conjunction with the Kinetic Materials Research Group, an origami tower with a Kresling folding pattern has been chosen as the current actuation vessel of the crawling robot, converting the rotary input of a servo motor into forward longitudinal locomotion. Two different variations of the crawling robot have been prototyped. The single tower robot is only capable of 1D motion and has been used extensively for the proof of concept testing for segmentation. As the segmentation concepts are finalized, they will be applied to a crawling robot with 2 internal towers, allowing for 2D locomotion. The robot with 2 internal towers has been prototyped in both a large scale and a small scale to demonstrate the scalability of the actuation approach, and segmentation will first be applied to the large-scale prototype.

A tiny Arduino and battery allow the crawling robot to operate untethered and allow for semi-autonomy in the form of path following. The current segmentation method uses SMA wires controlled by the Arduino. A signal is applied to the wires which contract and rotate a disk, shearing apart and deactivating the magnets connecting the segments. The feet of the robot have anisotropic friction due to their multi material design. The feet pivot on a pin joint, allowing the low friction material part of the foot to slide as the robot is moving forward and the high friction material part of the foot to keep the robot from moving backwards. To validate the crawling robot design, an analytical model is being developed to characterize the forward locomotion of the robot as a function of the input rotation supplied by the servo motor.

From the existing robot prototypes, the project is moving forward by increasing the autonomy of the robot to include path planning as well as path following. Improvements are also being made to increase the softness of the robot via exploration of new materials and reconsideration of the current actuation method.