Session: ASME Undergraduate Student Design Expo

Paper Number: 173842

Sensorized Humanoid Soft Hand 

The mechanical complexity of the human hand lies in its unique integration of muscle-driven actuation and tactile-sensing-enabled real-time feedback. This seamless interaction enables the dexterity required for gripping, positioning, lifting, and manipulating objects in everyday tasks. These capabilities depend on specialized nerve endings that provide sensations such as pressure, temperature, pain, and vibration. The sense of touch is essential not only for recognizing objects, but also for adjusting grip force and executing precise manipulations.

Building upon a prior prototype of a soft robotic hand with pneumatically actuated fingers, this work seeks to address limitations in both sensing and actuation. The original design incorporated a basic strain sensor in each finger to provide feedback. While the concept demonstrated feasibility, the strain sensors were limited in sensitivity, restricting their usefulness for capturing nuanced interaction data. To address this, the updated design embeds a flexible piezoresistive sensor array in each finger. These sensors provide higher resolution feedback, enabling detection of object properties such as distributed pressure and fragility with greater precision.

To develop this improved system, extensive CAD modeling and design iterations were conducted, informed by prior work in soft robotics and studies of human-inspired manipulation. Prototypes were fabricated using elastomeric casting and pneumatic assembly, and sensor performance was characterized through experimental testing and data analysis. This combined computational and experimental approach allowed both structural improvements and sensing upgrades to be systematically evaluated.

The effectiveness of sensing in the prior design was further constrained by the limitations of its pneumatic fingers. With a maximum grip strength of roughly 100 g, the hand could not manipulate heavier or irregularly shaped objects, resulting in fewer opportunities for meaningful data collection. To overcome this, we introduce auxiliary structures around the wrist that replicate the biomechanical role of human forearm muscles. In the human hand, forearm muscles and their tendons generate much of the grip strength required for object manipulation. Inspired by this principle, the new design employs a hybrid actuation system that integrates the existing pneumatic finger actuation with tendon-driven actuation anchored to soft, elastomeric pseudo-muscles in the wrist/forearm region. These pseudo-muscles, modeled after McKibben artificial muscles, are pneumatically actuated to augment grip force and broaden the range of motion.

The integration of enhanced tactile sensing with hybrid actuation substantially increases the robotic hand’s utility. Together, these improvements enable manipulation of a wider range of objects across variations in size, shape, weight, and material properties. Beyond advancing robotic manipulation, this work contributes to the field of soft robotics by demonstrating a bio-inspired, multi-modal approach to actuation and feedback. The resulting intelligent soft-robotic hand also presents promising opportunities for biomedical applications, particularly in the development of prosthetic hands that better replicate both the structure and functional nuances of the human hand. By combining improved sensing resolution with increased grip strength, this design offers a more complete platform for studying human-inspired manipulation and advancing assistive robotic technologies.

Presenting Author: Ariadna Ramirez University of Arkansas

Presenting Author Biography: Ariadna Ramirez-Lazaro is an undergraduate Mechanical Engineering student and Honors College member at the University of Arkansas. She conducts research as a Research Assistant in the Wan Research Group under Dr. Wan Shou, contributing to multiple projects in soft robotics and flexible sensor development. Her work spans bio-inspired soft robotic hands, millimeter-scale soft robots, and industry-focused design projects, with a current interest in exploring flexible humidity sensors. This research integrates experimental design, CAD modeling, and sensor characterization to advance robotic manipulation, sensing technologies, and biomedical applications.

Originally from Peru, Ariadna discovered her interest in STEM and bioastronautics—the study of the physical and psychological effects of space travel on biological systems—during a gap year exploring diverse projects in engineering, education, and art. She has been recognized with an Honors College Research Grant for her work, which combines innovative robotic systems with applications ranging from prosthetics to space technologies.

Authors:

Ariadna Ramirez University of Arkansas
George Vessey University of Arkansas
Wan Shou University of Arkansas