Testing and Optimization of High Q-Factor Magnetoelectric Devices
Biomedical Engineering and Computer Science, University of Minnesota
Advisor: Troy Olsson
Mentor: Sydney Sofronici
Biomagnetic sensing requires not only high device sensitivity, but also a large bandwidth (∼1kHz). MEMS devices utilizing magnetostrictive and piezoelectric materials show promise in these applications. Additionally, a low frequency feature of these devices allows for use in Wireless Power Transfer (WPT) applications. Optimal performance is achieved by driving the device at resonance under a DC magnetic bias field. Here, a testing configuration was designed and assembled to accommodate these features. To maximize Signal to Noise Ratio (SNR), the DC bias field was tested at various magnitudes and angles. A system of two sets of electromagnetic coils were arranged orthogonally to achieve a rotating DC field. SNR and Q-factor data from the device is reported for the higher (∼6.5MHz) and lower (∼80kHz) frequency modes in varied DC bias fields. Future steps include integrating permanent magnets at the appropriate angle into the device board design as well as the use of Hall magnetic sensors to continuously characterize this bias field.
Optimized Encoding of Sensory Stimulation for Brain-Machine Interfaces
Neuroscience and Electrical Engineering, University of Washington
Advisor: Andrew Richardson
Restoration of sensory feedback in Brain-Machine Interfaces (BMI) currently relies on user-in-the-loop calibrations or closed-loop approaches to replicate known neural responses corresponding to the target sensory perception. However, both approaches are not practically robust when considering the variability among users and neural interfaces. We propose an alternative, robust, and generalizable stimulus parameter search algorithm to 1) reduce the number of trials and total tuning time required to find discriminable sensory percepts, and 2) diversify the exploration of the stimulation parameter space.
Engineering Grasp Versatility Through Static and Active Stiffness Control of Robotic Grippers
Mechanical Engineering, Yale University
Advisor: Kevin Turner
Mentors: David Levine and Christopher Stabile
Traditionally, robotic grippers utilize rigid materials that enable the grasping of heavy objects but can cause damage to fragile objects. In contrast, while soft materials allow for safe interactions with fragile objects, they lack the stiffness required to handle large loads. For versatile grasping, an ideal gripper should be able to handle both heavy and delicate objects. Thus, a material system with a stiffness that can be controlled in real time is an attractive solution for achieving grippers with increased versatility. Here, we propose a soft gripper with a stiffness that can be switched from a soft state to a stiff state, allowing for grasping of a wider range of objects. By attaching electroadhesive clutches to the length of the gripper fingers, we can switch between high and low bending stiffness to handle heavy and delicate objects, respectively. We designed a testbed with controllable vertical and horizontal degrees of freedom to conduct grasping experiments, and we examine the effect of static and active stiffness change on load capacity using a closed-loop system enabled by the custom testing setup.
Project Report | Presentation Slides | Poster
Investigating the Capabilities of Miniature Autonomous Surface Vehicles through a Game of Pong
Computer Engineering, University of Maryland, Baltimore County
Advisor: Ani Hsieh
Monitoring underwater weather is important to understanding and maintaining the health of large water bodies, such as rivers, lakes, and oceans. Leveraging autonomous surface vehicles (ASVs) with on-board sensing capabilities can provide more useful and consistent information that captures the state of these water bodies. While large ASVs are currently in development at the ScalAR Lab, we simulate the performances of similar, yet smaller boats, miniature ASVs (mASVs). Investigating the performance of planning algorithms on board these mASVs can allow for a deeper understanding of the capabilities of coordinated tasks with teams of ASVs. In this project, a real-time planning algorithm for the mASVs is developed in Python and implemented to run in the Multi-Robot Tank (MR tank), to simulate a game similar to pong. In this game, miniature boats deflect an object between one another with a given set of boundaries. With the development of these planning algorithms, these concepts can be implemented into completing tasks for teams of larger ASVs. The implications may include the coordination of these vehicles in completing their given objectives, such as transferring waste materials collected from the surface of the river between multiple ASVs.
Design, Model, and Application of an Electromechanical Variable Stiffness Spring
Robotics Engineering, Worcester Polytechnic Institute
Advisor: Cynthia Sung
This paper outlines a design for a spring that can produce large changes in stiffness in a short period of time, while also maintaining other spring parameters such as relaxed length or diameter. To achieve this, we propose an electromechanically tunable stiffness spring. This mechanism consists of a thin PET sheet bent into a ring with many layers that we can add or remove continuously from the center to change the effective thickness of the ring without affecting its outside diameter. We built a model, and validation using a multitude of practical stiffness tests while changing spring parameters. Finally we demonstrate the capability of our spring in a tunable stiffness manipulator. By using our deformation model, we build a variable stiffness spring segment with 4DOF for millimeter accurate positioning, with rigidity control.
Understanding the Relationship Between Baxter Robot Approach to Contact Parameters, Perception of Safety, and Contact Dynamics
Mechanical Engineering, Worcester Polytechnic Institute
Advisor: Michelle Johnson
It is expected that there will be a rise in stroke patients without a sufficient rise in care professionals and facilities for rehabilitation. Robot-assisted therapy is one solution, but it must be conducted safely. This project analyzed how speed relates to perception of safety when a robot is acting as a physical therapist assisting a patient with a task. The project also showed patterns associated with contact dynamics using ‘lossy’ fiber optic force sensors. These identified how the Baxter robot recognizes when it has made contact with a patient and informed how to make therapeutic human-robot interactions be perceived as safe.
Engineering Quantum Biosensors with Nanodiamonds
Electrical Engineering, Penn State University
Advisor: Lee Bassett
Mentor: Henry Shulevitz
Because of their size, colloidal stability, room temperature quantum properties, and exceptional biocompatibility, nanodiamonds (ND) containing nitrogen-vacancy (NV) centers have many promising applications in quantum sensing, biolabeling, and nanomedicine. By conjugating nanodiamonds to various biomolecules, such as antibodies, therapeutics, and reporter particles, we can create a nano-sensing platform with simultaneous detection, diagnostic, and treatment capabilities. However, current methods of conjugating biomolecules to the nanodiamond surface have low yield and a high tendency of forming agglomerates. Here, we investigate ways to prevent nanodiamond agglomeration and increase conjugation yield by nanodiamond emulsions to increase the number of potential conjugation sites by a factor of 4 X.
Voltage Controllable Visible Light Range Reflector Display
Electrical and Computer Engineering, University of Illinois at Urbana-Champaign
Advisor: Marc Miskin
Technology is always evolving to be more reliable, user-friendly, and easy to produce. Developing screen technology in smartphones and tablets using electrophoretic display (EPD) screens instead of Liquid Crystal Display (LCD) screens would lower costs, weight, power consumption, and improve user experience. A downside of EPDs is that they are only reliable in black and white technology. This problem is approached by developing a fabrication process for colored e-ink actuators and voltage probing of the actuators. To achieve reliable and consistent actuators, different actuator designs were developed.
Characterizing Tissue Damage and Optimizing Material Properties of Tape Spring Based Steerable Needles
Electrical Engineering, Princeton University
Advisors: Mark Yim, Omar Abdoun
Often physicians must deliver treatment or take biopsies from difficult to access regions of the body. With traditional straight-shaft needle designs, tough tissue areas and bones blocking the needle’s path can complicate the process. Nontraditional steerable needle prototypes have thus been proposed as potential solutions to this problem, providing physicians more flexibility as they steer the needle to a precise treatment location. This paper outlines the material modifications made to a novel tape spring steerable needle prototype that allows for the design of a steerable needle that minimizes the necessary insertion force into tissue through a pointed, sharp-tipped tapespring based needle with vibrational capabilities. In the future, this improved needle prototype will be refined by assessing its performance when inserted in real tissue of nonhomogeneous stiffness, similar to the conditions such needles would face in the human body.
Wearable MXene Bioelectronics for Achilles Tendinopathy Diagnostics and Rehabilitation
Biomedical Engineering, The University of Texas at Austin
Advisors: Raghav Garg, Flavia Vitale
The Achilles tendon plays a crucial role in enabling multiple types of dynamic movement; however, as one of the most fragile structures in the body, its injury is extremely common. Acting in conjunction with multiple muscles, successful recovery of the Achilles tendon requires continuous monitoring of the loads that these muscles are imparting on it to avoid further detriment. To provide this insight, a novel application of MXene-based high-density surface electromyography (HD-sEMG) was innovated to capture muscle activation at specific points in the posterior lower leg. Spatial maps were generated to compare muscle activation across various activities and infer the resulting stress being placed on the Achilles tendon. The comfort, customizability, and high detail afforded by the investigated arrays propounds their future application in restorative physical therapy for tendinopathy patients, as well as in healing processes of other muscle groups and injuries.