2024 Design and Research Conference

Smart Load System

Team Members:  Skyler McAffry, Lindsey Miller

Advisors: Dr. Matthew Hartmann and Dr. Sandra Zivonavic

The Smart Load System is a small-scale smart breaker panel that distributes 120 Vrms to various circuits based on user-specified priorities. When used in residential applications, this system could be implemented to save money by avoiding the use of high-load circuits, such as a water heater, electric vehicle charging stations, etc., during peak load times. This is accomplished by tracking typical power usage within the system and having adjustable settings to allow the user to make changes as needed. A touchscreen display is used to adjust the priority level of each circuit, and a timer is used to delay when each circuit receives power. This touchscreen display also shows the current and average power drawn by each circuit and the voltage across the system. The system is protected using fuses of various sizes and pre-set logic within the PLC to limit the total amount of current used by the entire system to 15 Arms. This complies with the NEC code Section 210.21(B)(3).

PowUrgency

Team Members: Moaid Aldlaigan, Jake Bodie, Dylan Guillory, Jace Warren

Advisors: Dr. Matthew Hartmann and Dr. Sandra Zivonavic

The PowUrgency is an emergency power solution that provides a blend of reliability, sustainability, and efficiency. The name “PowUrgency” encapsulates the essence of urgency in providing dependable power during critical moments. Applications of PowUrgency spawn across various scenarios, from powering small household appliances during blackouts to providing essential energy in off-grid locations. Unlike its counterparts, PowUrgency stands out for its environmental friendliness, offering a sustainable alternative to traditional fuel-powered generators. By utilizing a 12 V deep-cycle battery and modern DC to AC inversion methods, the design ensures seamless operation of household electronics, including but not limited to light bulbs, phone chargers, and TVs. The system is designed to be capable of running at least 600 W load for extended periods. Carefully selected components including advanced safety features like low-battery cutoff and cooling fan circuits guarantee the long-term durability of the design.

TrafficBot

Team Members: Drew Guidry, Bel Panharith

Advisors: Dr. Matthew Hartmann and Dr. Sandra Zivonavic

The TrafficBot: Single Lane Closure Traffic Control Device Using Machine Learning is a system that uses images of cars to learn what cars look like, then uses this information to detect cars in a single lane closure and direct them safely and efficiently through the closure. The two 12V batteries power our system and uses a combination of microcontrollers and cameras to collect and use data. Then, once our system detects cars, it will use two 12V motors to turn a sign that directs cars to move slowly through the lane closure while the other side is stopped. Our system effectively will work the same way as having two people holding signs, but instead of having people turning signs, the system will run automatically, making it much safer than people standing on the road. Our design is a safe, innovative approach to directing traffic around construction zones, car accidents, and other road hazards.

BREAK

Unbalanced Load Power Meter

Team Members:  John Burrell, Jacob Cain, Bradley Reed

Sponsor: Dr. Prashanna Bhattarai

Advisors: Dr. Matthew Hartmann and Dr. Sandra Zivonavic

Unbalanced loads within three-phase systems introduce inaccuracies to power calculations, which are not accounted for by conventional power meters. Such inaccuracies provide incomplete information on the power losses, power rating, and power supply in unbalanced three-phase systems. Our device, the Unbalanced Load Power Meter, utilizes the “Currents’ Physical Components” (CPC) concept to create an algorithm which calculates and displays more accurate apparent power measurements. This algorithm accounts for unbalanced loads present in a system, leading to a higher degree of accuracy than meters which neglect the unbalanced component of apparent power. Utility companies and factories with high-power equipment would be the primary beneficiaries of this product, as small inaccuracies in apparent power metering can lead to large aggregate losses in energy, equating to an increase in operational costs. To maintain quality throughout testing, we specified several metrics to guide our design: operating within a 3-phase range of 120 and 5 in both wye and delta configurations; providing an output with an accuracy within 1% error; and possessing a safe, lightweight, and portable housing

Beamforming via Reconfigurable Intelligent Surfaces

Team Members: Charles Dalferes, Jared Matherne, Nathan Scheffe

Sponsor: Dr. Jinyuan Chen

Advisors: Dr. Matthew Hartmann and Dr. Sandra Zivonavic

With the ever-increasing number of smart devices that use wireless communication to constantly send and receive signals/data, the current systems in place are going to eventually falter in the efforts to maintain the mass traffic. The 6G technology is expected to utilize millimeter waves, which have the capacity to transmit a large amount of data but are more susceptible to noise and occlusions. A Reconfigurable Intelligent Surface (RIS) utilizes beamforming to transmit a signal with very little noise in the received signal. Beamforming is a technique that uses multiple antennas placed very close to each other to transmit multiple signals that are identical except for their phase. These signals constructively interfere at the targeted area, resulting in a high signal-to-noise ratio. Our team designed a 2-bit modular unit cell that is capable of emulating four different patch antennas. These patch antennas use PIN diodes to open/close the connections to extra patches, changing the affected surface area and adjusting the reflection coefficients to fit our need. This process is automated using a microcontroller with a hierarchical codebook. The performance of the unit cell is scalable, which allows us to create a larger array to increase the total dB gain.

Water Quality Detection System

Team Members: Raijin Bolden, Sawyer Kees, Noah Laughlin, Logan Tharp

Advisors: Dr. Matthew Hartmann and Dr. Sandra Zivonavic

Our project focuses on providing multiple water quality indicators to ensure safe drinking water. These sensors include the pH sensor, which should accurately measure 6.5 – 8.5 (∓ 0.1), The TDS probes, which help measure the conductivity value and should be below 1,000 (µS/cm), The ORP should float around the 250 mV area without throwing a warning, and the turbidity sensor (should be below 1 NTU). The system, in totality, takes the advanced sensors to provide real-time monitoring, allowing swift decisions for public safety. Additionally, the goal for the chemical detection is to throw areas for elements that are too high, such as lead (below 0.015 mg/L), for safe consumption. By employing the help of some motion tracking cameras and test strips, we can determine some of the most important factors for water quality.

FRESH 3D Bioprinter

Team Members:  Evan Goodman-Blue, Nicholas Jones, Timothy Searcy, Sarah Sellen

Sponsor: IEEE Nanotechnology Council

Advisors: Dr. Sandra Zivanovic, Dr. Matthew Hartmann and Dr. Adarsh Radadia

For this project, we prototyped a low-cost multi-material 3D bioprinter, utilizing the single-material FRESH 3D bioprinter developed by the 2022-2023 Nanosystems Engineering senior design group. Our team improved the Peltier cooling system, redesigned and fabricated the exterior shell of the printer, and rebuilt the nozzle assembly to allow for multi-material printing, all while maintaining the system’s ability to be used in low-gravity environments. These modifications improve the quality of the printed materials, allow for the creation of more complex structures, and make the overall design more commercialized.

Automatic Greenhouse

Team Members:  Coy Disher, Cheston Sturdivant, Jace Ziegler

Advisors: Dr. Sandra Zivanovic, Dr. Matthew Hartmann and Mr. Maxwell Hide

The Automatic Greenhouse is a small, indoor greenhouse control system comprising four main subsystems: lighting, temperature, humidity, and soil moisture control. An Arduino automatically controls these four subsystems to optimize growing conditions within the greenhouse. A full-spectrum grow bulb supplements light. Temperature and humidity are controlled by two fans, a heater, and vents in the greenhouse. The watering system allows up to three potted plants to be watered separately within the greenhouse. There is a plant library with preset growing conditions from which users can choose. Users can manually input preferred growing conditions for plants not in the library.

Printed Circuit Board Stator for Multirotor UAV Flight

Team Members:  Luke Fussell, Ian Golsby, Brandon Oubre, Caleb Reichard

Advisors: Dr. Sandra Zivanovic and Dr. Matthew Hartmann

Modern advances in printed circuit board (PCB) manufacturers have facilitated the creation of a fully functional uncrewed aerial vehicle (UAV) on a monolithic PCB. Additionally, the PCB motor industry has a market gap for medium-torque output and high angular velocity motors, which this project addresses. Direct integration of a PCB motor into a PCB UAV yields enhanced efficiency and reduced complexity. This project’s team designed and implemented axial flux permanent magnet (AFPM) motors utilizing a PCB stator into a fully PCB drone to enable flight. The drone uses four novel motors to produce thrust. The custom motors use 16 disk-shaped magnets in a 3D-printed rotor structured in a single-stator double-rotor AFPM configuration. The control of the drone is enabled through an onboard, custom flight controller and four electronic speed controllers (ESC), all of which were designed in KiCad. The systems implemented in the flight controller include an inertial measurement unit (IMU), switching-mode power regulation, and a microprocessor.

BREAK

Marlin: Object Recognition Enabled Submersible Vehicle

Team Members: John Able, Alexis McCarthy, Sonny Reed

Advisors: Dr. Sandra Zivanovic, Dr. Matthew Hartmann and Dr. Jinyuan Chen

In the United States, dive teams everywhere are deployed for underwater crime scene investigation and evidence recovery. Also, dive teams aid in the maintenance and upkeep of underwater structures like bridges and oil rigs. This underwater recovery or maintenance method is time-consuming, expensive, and dangerous for all dive team members. This is the problem that our project, Marlin, intends to solve. Marlin is an object recognition-enabled submersible remote-operated vehicle that can aid in discovering objects of interest under the surface of water. The objects of interest can be anything from a crack in an oil rig’s pipe to a piece of evidence in a criminal case discarded into a body of water. While defining this project’s scope, we conducted research that dives into machine learning-based algorithms and AI and methods of creating a DIY remote-operated vehicle for exploration-based purposes. Using this research, we created a proof-of-concept prototype using four brushless motors and a small web camera connected to a battery pack that solves the problem.

Carbon Quantum Dot Heavy Metal Sensor

Team Members:  Yashadara Ekanayaka, Jared Melseth, Jonathan Tairov

Sponsor: IEEE Nanotechnology Council

Advisors: Dr. Sandra Zivanovic, Dr. Matthew Hartmann and Dr. Shengnian Wang

Heavy metal pollution poses a significant risk to our health and environment, as heavy metals are very toxic and highly resilient. Therefore, tracking their presence in the environment, such as water or soil, is essential. However, traditional methods of identifying these kinds of contaminants require costly and specialized equipment. Carbon quantum dots (CQDs) pose a cost-effective method of heavy metal detection by utilizing their fluorescent properties. In this project, we investigated the effects of copper pollutants on CQDs’ fluorescence. We synthesized CQDs by a one-pot hydrothermal reaction of citric acid and ethylenediamine. We were able to measure and mathematically correlate the fluorescence quenching from the addition of copper. We found that we could identify the concentration of copper within the 1-25 mM range. Alongside the design of a cost-effective chemical sensor, we maintained the accessibility of our system by creating a user-friendly Arduino experience that enables accurate sensing for non-technical people. Access to pollutant sensing technology could enable people worldwide to make more informed decisions to help manage ecological disasters before they become widespread.