Search

Electromagnetic actuators provide fast speed, large forces, high strokes, and wide bandwidths. Most designs, however, are constructed from rigid components, making these benefits inaccessible for many soft robotics applications. In this work, we develop a new soft electromagnetic linear actuator using liquid gallium–indium for the conductor and neodymium–iron–boron and polymer composites for the permanent magnet. When combined in a solenoid configuration, high strokes can be generated using entirely soft components. To emulate the pulsing motion of Xenia coral arms, we develop an additional soft flexure system that converts the linear translation to rotary motion. The design and fabrication of the electromagnetic actuator and compliant flexure are first described. Models for the magnetic forces and the joint kinematics are then developed and compared with the experimental results. Finally, the robot dynamics are analyzed using stochastic system identification techniques. Results show that the compliant actuator is able to achieve an 18 mm stroke, allowing the soft arms to bend up to 120 deg. This further enables the tips of the arms to traverse an arc length of 42 mm. Bandwidths up to 30 Hz were also observed. While this article focuses on emulating a biological system, this highly deformable actuator design can also be utilized for fully soft grasping or wearables applications.

Unique tags are often added to items to prevent counterfeiting and tampering. While there are many techniques for generating unique patterns, tags that are inexpensive, scalable, easy to read, and difficult to replicate are particularly useful for tracking applications. In this work, we present a novel tagging concept that utilizes self-assembly methods to create physically unclonable patterns using magnetic particles. By combining a distribution of Nd-Fe-B magnetic particles with a hard polymer inside a strong magnetic field, patterns are generated that are easily read using magneto-optic techniques. Image processing and template matching techniques are then used to determine the likelihood that two different magnetic tags are classified as the same. For optimal selections of particle percent mass, statistical analysis predicts that pattern matching yields at most one false positive for every 1015 true positives. Results further illustrate that variations in particle size and particle percent mass have a strong effect on the predicted pattern uniqueness. Finally, a magnetic field model was then used to help predict broad trends in pattern statistics and particle structures. This paper shows that magnetic self-assembly patterns hold promise as unique embedded tags for a large range of applications from microelectronics to luxury goods to low-cost off-the-shelf items.