Academy of Interdisciplinary Studies Division of Integrative Systems and Design 253 Mechanically Intelligent Gripper for Fruit Pickup Supervisor: Rob SCHARFF / ISD Student: WEI Yang / ISD Course: UROP 1100, Spring This study presents a miniaturized octopus-inspired suction cup designed for soft continuum robotic arms, integrating advanced adhesion and sensory capabilities. Modeled after Octopus vulgaris, the suction cup features a dual-chamber structure and ChromaTouch markers for real-time adhesive strength estimation. Parametric modeling with 34 adjustable parameters enables rapid customization, while PolyJet 3D printing ensures precise fabrication using Agilius materials. An endoscope camera reduces size without compromising performance, and a single-pump system enhances efficiency for multi-cup applications. This design advances soft robotics by enabling robust manipulation and perception in unstructured environments. Future work focuses on integrating multiple suction cups for enhanced soft arm functionality. Seahorse-Inspired Actuator for Underwater Manipulation Supervisor: Rob SCHARFF / ISD Student: WONG Lai Yin Garmisch / ISD Course: UROP 1100, Fall This project focuses on developing and characterizing a 3D printed seahorse tail inspired soft actuator. The soft actuator developed is pre-curled in the unpowered state and would straighten and unwind when negative pressure is provided; curls up when positive pressure is provided. This novel approach is unlike previous works where those actuators would straighten when positive pressure is supplied (Antonelli et al., 2024). This allows our actuator to generate a much larger hooking and gripping force (79N). To study the relationship between the geometry of the actuator and its performance, we designed the model to be parametric, allowing us to easily generate actuators with different geometric parameters, such as a different logarithmic spiral, air chamber shape and density. Such that we can find the best performing actuator for different applications, such as underwater anchoring and gripping. Soft Robotic Fish for Underwater Exploration Supervisor: Rob SCHARFF / ISD Student: LIU Hualin / ISD Course: UROP 1100, Spring UROP 2100, Summer This phase of the project advances the GRACE pneumatic artificial muscle (PAM) toward closed-loop control by completing a comprehensive testbench, calibrating all sensing modalities, and establishing robust mappings from physical quantities to ADC codes on an STM32G431 microcontroller, The platform integrates a proportional valve, air pressure sensor, force sensor, distance sensor, and embedded optical fiber intensity sensing, with synchronized acquisition and logging. We detail the mechanical, pneumatic, electronic, and software architecture; sensor calibration procedures; quantization-aware conversion pipelines; and data acquisition protocol for generating training datasets. Initial trials confirm stable actuation and repeatable sensor responses with low drift over typical experiment durations. Ongoing work focuses on systematic dataset collection across randomized pressures and discrete muscle lengths to train and validate machine-learning models for proprioception and closed-loop control.
RkJQdWJsaXNoZXIy NDk5Njg=