STEM I Project

Short Course Description

STEM with Science and Technical Writing is taught by Dr. Crowthers. In STEM, students develop Independent Research projects and complete lots of scientific and technical writing.

STEM Title

QR Code

Abstract

Soft robotics offers significant potential in the field of surgical robotics. Soft materials, due to their high adaptability, deformability, and number of degrees of freedom, often mimic the mechanical properties of biological tissues. This makes them ideal for use with delicate tissue and in several different surgical applications, but often due to their inherent properties, it becomes challenging to receive feedback on soft tools being used in the surgical field. This project aims to develop a smart touch system for surgical robots that enhances real-time feedback by sensing force and torque during minimally invasive procedures. Addressing the limitations of current robotic systems in detecting nuanced tissue properties, the proposed design incorporates hybrid soft robotics and deep learning to classify tissue type, texture, and applied force. By integrating force sensors, piezoelectric sensors, soft actuators, and feedback loops, the system will provide surgeons with precise tactile insights. To achieve this, a proof of concept was developed, integrating a load cell into an end-effector CADed using Onshape. Expected outcomes include a functional prototype demonstrating accurate force and texture detection, evaluated using benchmark tissue models. The design will feature two end-effectors equipped with force sensors and soft, flexible materials to mimic human touch sensitivity. Success will be assessed by comparing the system’s feedback accuracy and reliability to standard surgical tools. Simulations and iterative testing will ensure robustness before moving to physical validation. This research addresses critical gaps in surgical robotics by improving dexterity and sensory feedback, potentially reducing complications during surgery.

Research Proposal

Click here to view my research proposal!

Engineering Goal

The objective of this project is to develop a controllable real-time force sensing system for minimally invasive surgery. By integrating sensors, the system will accurately measure force and applied to actuators, providing feedback on tissue properties.

Engineering Problem

Minimally invasive surgery (MIS) often lacks the tactile feedback necessary for precise surgical maneuvers. Existing robotic systems provide visual guidance but fail to accurately sense and deliver information on tissue properties such as force, texture, and compliance. This limitation increases the risk of accidental tissue damage and complicates delicate procedures requiring fine motor control.

Background

Soft robotics is an emerging field with potential in automated and teleoperated systems, particularly in surgical applications. Soft materials, such as silicone, elastomers, and other polymeric substances, are characterized by their low stiffness and high deformability, mimicking the mechanical properties of biological tissues. These materials allow robotic tools to exhibit flexibility and adaptability in confined or irregular spaces and have higher degrees of freedom while reducing the risk of damage when interacting with delicate structures, such as human organs. Current methods to enhance robotic tactile sensing often integrate soft robotics with external tactile sensing technologies, such as electronic skins (e-skins) and electronic textiles (e-textiles) (Jahanshahi & Zhu, 2025). E-skins are thin, flexible materials embedded with sensors that mimic the sensory functions of human skin, capable of detecting pressure, temperature, and texture. E-textiles are fabrics integrated with conductive fibers and sensors to measure mechanical interactions such as force or stretch. These technologies improve the perception and measurement of forces applied by robotic tools, providing critical data for precise control. However, the challenge remains developing a comprehensive system capable of delivering precise and actionable haptic feedback to surgeons (Tiwana, 2012).This feedback is essential for distinguishing tissue types, assessing applied force, and making real-time surgical decisions with greater confidence and accuracy. Research in human-robot interaction emphasizes redundancy for smooth, safe, and adaptable robotic motion, as demonstrated by Li & Wang (2024), who explored the application of a three-dimensional force sensor in robots with extra degrees of freedom. A Jacobian matrix, which mathematically relates the velocities or forces at the end-effector of a robotic system to the joint velocities or forces, was utilized in this study to optimize motion performance. The augmented Jacobian matrix was specifically designed to maintain safety margins, avoid joint limits and obstacles, and enhance adaptability in dynamic environments. In this research, the outcome of the enhanced analysis was beneficial and although the sensor was implemented in a largely rigid model, it ended up being more adaptable and conformable to different scenarios. Furthermore, continuum robots, such as those discussed by Xu (2009), are robotic structures designed to mimic the flexible and continuous motion of biological systems. These robots achieve high dexterity and adaptability in confined spaces through their segmented or flexible designs, often composed of soft materials or elastic joints. While these systems offer significant advantages for minimally invasive surgeries, they face challenges in incorporating precise tactile sensing and force feedback mechanisms due to their reliance on soft actuators and the absence of rigid structures for traditional sensors. Similarly, hybrid manipulators for lower gastrointestinal interventions, as seen in recent advancements, highlight the importance of force estimation for handling delicate tissues but often encounter difficulty in maintaining accurate control and feedback in complex anatomical environments (Leung, F. et al., 2024). However, these systems often struggle with stability, sensitivity, and adaptability in dynamic environments. For instance, robotic catheter systems actuated with cables can have limitations such as low amounts of stability, high loss in force, and inadequate sensing of force, as noted in the development of a soft robotic catheter (Nguyen et al., 2023). These gaps highlight the need for innovations that enhance real-time tactile sensing, as surgeons often lose tactile perception while utilizing less direct instruments. There is also a need to provide better force feedback, and improve system stability without compromising dexterity. This project aims to develop a smart touch system for surgical robots by integrating intrinsic tactile sensors, force sensors, piezoelectric sensors, hybrid soft robotics, and feedback loops to help compile and get data. This approach will enable precise detection of tissue properties such as type, force, and texture, and while providing real-time feedback and data to enhance surgical precision and outcomes. By offering surgeons detailed tactile information, the system will mitigate the lack of sensory feedback in minimally invasive procedures, allowing for safer manipulation of delicate tissues and improved navigation in confined anatomical spaces. Specifically, the system will incorporate 3D-printed robotic end effectors with embedded sensors that analyze touch data using machine learning. These tactile cues will guide surgeons in making critical decisions during operations, reducing the risk of tissue damage and enhancing the overall efficiency of complex surgeries.

Procedure

Fluidic Control Board: • Integrated a 100 PSI gauge pressure sensor to monitor real-time pressure levels • Used 3-way normally closed solenoid valves (24V, high flow) to precisely control airflow • Incorporated opto-isolated power MOSFET switches for efficient valve actuation • Assembled a high-flow manifold to distribute air evenly across actuators • Employed miniature diaphragm air pumps for controlled pressure delivery • Installed toggle switches and potentiometers for manual control and calibration • Utilized a step-up/down voltage regulator (5-25V input, 0.5-25V output) to ensure stable power supply Soft Actuator Fabrication • Designed a 3D-printed mold using Onshape CAD for actuator casting • Manufactured actuator using Elastosil M4601 silicone elastomer for high flexibility and durability • Embedded pneumatic tubing (OD 1/8") within the actuator for controlled inflation • Sealed and reinforced actuator structure to prevent air leakage and optimize deformation Sensor Integration • Embedded force sensors within the actuator structure to measure real-time mechanical interactions • Connected sensors to an Arduino Mega, programmed for continuous data acquisition • Calibrated sensors using controlled weight applications to establish force-voltage relationships Testing and Data Collection • Evaluated actuator deformation and force response under different pressure levels • Conducted experiments using known weights (100g, 200g, 500g) to validate force sensor accuracy • Visualized sensor outputs using Arduino Serial Monitor and Serial Plotter for real-time data analysis • Tested actuator performance on synthetic models to assess surgical feasibility

Figures

Force v. Pressure Readings with time. Test was done without any force to determine the baseline effect of the pressure changes on the force sensor, which were insignificant. The force sensor was calibrated both inside and outside the actuator, with strong linear correlations in both cases (R² = 0.950 without actuator, R² = 0.966 when embedded).

Analysis

Discussion of experimental results.

Discussion/Conclusion

The results from Prototype 1 indicate a strong linear correlation between applied weight and sensor voltage, demonstrating the reliability of the piezoelectric sensor in detecting applied forces. Objective 1a and 1b were met, as the collected data from both force and torque sensors provided accurate readings that could be effectively compiled for real-time analysis. The statistical analysis, including a high R² value of 0.932 from the linear regression, supports the accuracy and consistency of the system. Potential limitations included the use of a single actuator design, which may not generalize well to different surgical tasks, and the limited range of applied forces tested. Additionally, the reliance on piezoelectric sensors alone may limit the scope of tactile feedback. Despite these limitations, the research advances existing work on MIS tactile feedback systems by integrating soft robotics with multimodal sensing. This project successfully achieved its objectives by demonstrating real-time force and torque sensing with accurate tactile feedback using soft robotic actuators and piezoelectric sensors. The methodology, from proof of concept to prototype development, provided reliable data supporting the system's feasibility for minimally invasive surgery. The results showed a strong correlation between applied force and sensor voltage, validating the sensor's accuracy. Despite challenges such as sensor drift and actuator variability, modifications ensured consistent performance. This research not only contributes to the advancement of tactile feedback systems in robotic surgery but also highlights opportunities for future innovations in sensor integration, soft materials, and closed-loop control systems.