The objective of this EArly-concept Grants for Exploratory Research (EAGER) project is to advance the field of magnetic soft robotics by demonstrating a proof-of-concept mechanism for independently actuating multiple robots with a single external magnetic field. The underlying principle is essentially a magnetically controlled "combination lock," that prevents the robot from deforming in response to a command unless the command is preceded by the correct lock combination. The soft robots under study have potential applications to biomedical devices, surgical tools, and manufacturing. This effort is expected to advance scientific understanding of magnetic interactions in magnetic materials and soft structures, advance the use of machine learning for soft robot design, and promote progress in the field of small-scale soft magnet robots. The goal of this project is to design, fabricate, and evaluate a novel modular magnetic actuation mechanism that operates at scales ranging from millimeters to centimeters. This mechanism integrates an external dynamic magnetic field with innovative structural designs for soft robots to achieve decoupled, multidirectional tuning within individual units, and selective actuation across large arrays of robots. Design methods based on machine learning will be used to guide the design process and ensure that each robot is actuated reliably and independently. Additively manufactured and microfabricated prototypes will be systematically characteriz