Student Design Projects
Undoubtedly, one of the most talked about aspects of the major is the 1/C Capstone Design projects. During the 1/C year students get to design, build and test a project of their choosing. Depending on your interests the project can be an entirely unique idea, or it can be an entry in a design competition of some type, or perhaps a research project with a faculty mentor.
We'll let you build just about anything you can dream up (as long as its safe and we have the parts). Most teams consist of 2-3 students. Check out some of memorable capstone projects from previous years.
SWAT-C: Autonomous Target Detection Using a Neural Network
This project focused on the application of modern technology to help a squad size element in combat. To obtain that goal this project uses computer vision to improve the squad’s situational awareness. Computer vision has been used on drones for over a decade and has been proven to be more effective than other methods due to the limited size and power of commercial and military drones.
Microrobot Control Manipulation
Robotics has surged to the forefront of biomedical and healthcare research, providing a novel approach to combatting disease and injury. In an attempt to utilize robots on, and in, the human body, developing control algorithms for robots on small scales (milli-, micro-, or nanometers) in fluid environments is popular in medical research. In this project, we considered a previously discovered control law which computes pulse width modulation (PWM) signals to send to magnetic coils that control a microrobot’s position. Previous work assumed a priori knowledge of the environment and regulated voltage signals, which achieved unstable results. We proposed a new Robust Integral Sign of the Error (RISE) control method to be implemented on the USNA microrobot lab setup to provide positioning performance in the presence of viscous forces and varying external magnetic forces.
Tornique Research and Modeling
The purpose of this project is to design, model, and test an autonomous tourniquet concept. Autonomy was the focus of the project due to its correlation to optimum pressure and user-friendly nature. The team explored the mathematical models of Tuncali and Hong-yun for arterial occlusion pressure and incorporated their findings into a MATLAB model. The final physical design of the tourniquet system was an augmented field tourniquet with an attached Mckibben muscle that was hydraulically actuated by a syringe pump.
Use of Machine Learning for the Diagnosing of Metastasis in Lymph Nodes
The goal of this project was to develop and then compare a shallow neural network and a deep convolutional neural network that take a data set of cellular images as inputs and output regions of interest where cancer is likely present. The accuracy of these algorithms were evaluated by the number of false negatives in each slide due to their unique importance in cancer diagnosis. Even though this project only concerns a specific type of diagnosis, the result of finding the optimal algorithm to minimize false negatives is applicable to other medical conditions that can be diagnosed through imagery including lumbar problems.
Self-Assembling Single Actuator Robots
This project aimed to develop a viable network of homogenous single actuator robot, capable functioning both independently and as a single connected unit. Each robot is capable of independent motion in two dimensions using a single actuator; when joined differential drive principles allow the units to maneuver as one. Key components of this project include implementing a reliable communication system between network nodes, developing an accurate navigation method, and designing a physical linkage to effectively join the two robots.
Interoperable UGV -UAV System for Tracking and VTOL
The purpose of this project was to demonstrate proof of concept of an interoperable air-ground unmanned vehicle system. This system allows the ground and aerial components to communicate in a manner to establish a tracking-following relationship that incorporates vertical take-off and landing (VTOL) capability.
3-D Printed Prosthetic Hand
This project is a low-cost, highly customizable 3-D printed prosthetic limb for trans-radial amputees. The hand is controlled via myoelectric signals. In addition, a haptic feedback system is integrated into the prosthetic so that the user can "feel" force applied to the fingers of the prosthetic via vibro-motors.
Helium Balloon Surveillance
The goal of this project was to develop an ad-hoc network of large helium balloons which each had a sensor platform containing a camera and GPS. These students used coding to merge multiple video streams into a single mosaic to provide persistent intelligence, surveillance, and reconnaissance (ISR) over a large area.
Haptic Feedback Suit
Warfighters in the field are constantly tasked with an overwhelming set of responsibilities and goals. As a solution to the land navigation problem, these midshipmen developed a suit with built-in haptic actuators that were set to indicate the desired direction a user is supposed to go based on a desired path. This haptic feedback suit could significantly reduce the amount of time spent looking at a map and improve situational awareness. These midshipmen were the winners of the Marsh Award- the departmental prize for the best 1/C design project.
Lacrosse Ball Returner
These midshipman designed a device that could reduce time wasted retrieving shot lacrosse balls. They developed a cost-effective system capable of returning a lacrosse ball to a shooter's exact location while practicing shooting on goal.