麻豆传媒社区

This project explores a timely and important question: Can artificial intelligence (AI) be guided to recognize when structural damage is truly dangerous?

While many existing AI tools are designed to detect cracks in infrastructure such as bridges and buildings, very few focus on evaluating how serious that damage actually is. Engineers make this decision regularly by assessing whether damage is just on the surface or an early warning sign of failure. In Georgia and across the country, infrastructure is often vulnerable due to age, harsh weather, and heavy use. As a result, early identification of high-risk damage can play a critical role in public safety and cost-effective maintenance. This project aims to develop an AI algorithm, particularly computer vision, that can classify structural damage not just as present or absent, but as either minor or potentially dangerous, based on visual characteristics that engineers use during inspections.

Student(s) will actively participate in every phase of the project, regardless of their prior experience. They will train and test a computer vision model capable of recognizing damage severity. The faculty mentor will provide step-by-step training on all technical tasks, including data preparation, data processing, model development, model testing, and performance evaluation. Student(s) will also engage in reflection and discussion about the ethical implications of using AI in high-stakes decisions, such as what happens if an AI model underestimates the severity of a structural flaw. All activities will be supported through interactive tutorials and collaborative tools such as Google Colab. Final results will include a trained model, visual analysis, and a student-led presentation at the KSU Symposium of Student Scholars. For first-year students interested in engineering, technology, or the role of AI in protecting public infrastructure, this project offers an exciting opportunity to contribute to meaningful, real-world research in a highly supported and engaging environment.

The student(s) participating in this project will gain practical experience in the application of AI to challenges in structural safety. With step-by-step guidance from the faculty mentor, the student(s) will learn how to prepare and process database, train AI models, optimize hyperparameters of the model, test the developed model, comprehensively evaluate performance of the model, and report the results. This will include exposure to basic concepts in machine learning and computer vision. The student(s) will also learn how to interpret the model鈥檚 outputs and identify when and why the AI may struggle to correctly classify certain types of structural damage. These technical experiences will help build a foundational understanding of how AI tools can support real-world engineering tasks.

Beyond technical skills, the student(s) will engage in reflective discussions about the ethical and practical implications of using AI in situations where safety is involved. The student(s) will also develop communication skills by preparing visual summaries and explanations of their work, which will be shared through a poster presentation at the KSU Symposium of Student Scholars. Throughout the project, emphasis will be placed on problem-solving, critical thinking, and clear communication. No prior experience in programming or engineering is required. By the end of the project, the student(s) will have completed a guided research experience that contributes to an emerging area of study and provides a strong foundation for future academic or professional opportunities in engineering, computer science, or related fields.

Each week, the student(s) will engage in a series of guided activities designed to build understanding and skills progressively throughout the semester. In the early weeks, the focus will be on learning the background of the project, including how engineers assess structural damage and how AI can support that process. The student(s) will review a set of pre-labeled images of damaged structures and begin working with datasets that will be used to train and test a computer vision model. With the support of the faculty mentor, the student(s) will be introduced to the tools and platforms needed for the project, including step-by-step tutorials in Google Colab. Weekly meetings will provide opportunities to review progress, ask questions, and discuss the purpose and meaning behind each task.

As the project advances, weekly duties will include training the AI model, testing its ability to evaluate new images, and interpreting how well the model performs. In the final two weeks of the project, attention will shift to summarizing findings, creating visual materials, and preparing a poster presentation for the KSU Symposium of Student Scholars. All tasks are designed to be manageable within 5 to 10 hours per week and are structured to ensure continuous learning without requiring any prior experience in AI or engineering. The activities are primarily virtual and can be completed with Internet access.

Online

Dr. Tadesse Wakjira, twakjira@kennesaw.edu 

While concrete is essential construction materials worldwide, the production of concrete is a significant driver of global CO鈧� emissions, primarily due to the energy-intensive manufacture of cement and the extensive extraction of already exploited natural aggregates (sand and stone). This project investigates the development of super-recycled concrete, which incorporates ultra-high volumes of industrial waste including fly ash, recycled glass, plastics, and rubber powder as partial replacements for cement or fine aggregates. Unlike conventional designs for recycled concrete that focus on substituting coarse aggregates with crushed concrete, this study targets the cement鈥揻ine phase, where material interactions critically influence strength and durability.

By intervening at this microstructural level, we aim to engineer high-performance, eco-efficient concrete. The experimental phase will involve the design and testing of multiple recycled concrete mixes, with more than 50% replacement of traditional binders or sands using industrial waste products. Key performance indicators will include compressive strength, slump, air content, and electrical resistivity in pores, all evaluated through standardized laboratory procedures. To assess broader impacts, we will conduct a comprehensive Life Cycle Assessment (LCA) and Life Cycle Costing (LCC) analysis using an open platform. These will quantify the environmental benefits as well as the long-term economic viability of each mix design compared to conventional concrete.

This research aims to: 1) Develop concrete mix designs with 鈮�50% substitution of cement or fines using industrial waste; 2) Evaluate mechanical and durability performance through lab testing; 3) Contribute to the materials database developed by PI鈥檚 research group; 4) Quantify environmental impacts using ISO-compliant LCA methodologies; and 5) Analyze economic performance across the material's life span in a case study.

1. Gain experience in conducting scientific investigations, literature reviews, and data analyses
2. Acquire specialized knowledge in construction materials, specifically cement concrete mixes
3. Learn laboratory techniques and data analysis methods relevant to the proposed research.
4. Develop analytic skills in LCA and LCC through a real-world case study.
5. Present their work at conferences, contribute to materials database, publish the results in academic journals. 

1. Attend weekly progress meeting.
2. Prepare laboratory samples using various industrial waste materials and conduct mechanical testing under the supervision of PI.
3. Perform economic and sustainability analysis and modeling for super recycled concrete 
4. Assist in literature reviews and data visualization and prepare materials for posters or conference presentations.

Face-to-Face

Dr. Youngguk Seo, yseo2@kennesaw.edu 

Post Covid019, traffic congestion patterns have changed, specifically near elementary and middle school locations. More parents are choosing to use a personal vehicle as a primary mode of transportation to pick-up and drop-off their kids. In addition, parents of elementary and middle school kids who are part of the school choice program in Georgia must use the car to pick-up and drop-off their children. This process creates traffic congestion at intersections near schools during pick-up and drop-off times. 

The goal of this project is to develop a safe and efficient transportation solution for GA parents of elementary and middle school children. To achieve this goal, we will focus on two completing two objectives; (1) optimize traffic congestion at intersections near elementary and middle schools and (2) reduce total travel time for GA parents by developing a cloud-based application that integrates a smart-phone app with static vehicle detection sensors. Each objective will be accomplished by (1) understanding current state-of-the-art (i.e. what are current solutions being offered for the same problem); (2) identifying various possible solutions (i.e. brainstorming possible solutions and listing pros and cons of each solution); (3) developing the best possible solution (i.e. collecting data, developing traffic simulation models, or developing cloud-based applications).

• Understand the research process. 
• Learn how to conduct a literature review.
• Learn how to identify possible data collection sources online.
• Learn how to collect data in the field with data collection devices. 
• Learn basics of traffic flow and traffic signal operations.
• Learn how to develop and calibrate traffic simulation models.
• Learn how to develop and test cloud-based applications.
• Learn how to develop surveys for data collection.

Students will work with graduate students to process data and finalize Synchro network for selected elementary school. Part of the network has been developed in the past by previous research team. 

Synchro related tasks will require students to work in the G233 lab. Some of the tasks can be completed remotely. 

Hybrid

Dr. Parth Bhavsar, pbhavsar@kennesaw.edu 

Water plays a central role in our everyday lives鈥攊t flows through pipes to reach our homes, travels through treatment plants to make it safe to drink, and circulates through tanks and reservoirs to ensure a steady supply. Understanding how water moves in these systems is critical for designing safe, reliable, and efficient infrastructure.

This project introduces students to Computational Fluid Dynamics (CFD), a computer-based method that allows us to visualize and analyze how fluids (like water) move without building expensive physical models. With CFD, engineers can 鈥渟ee鈥� inside a pipe, tank, or reservoir to examine how water flows, where energy losses occur, and how design changes (like adding a bend in a pipe or a baffle in a tank) affect performance.

As first-year students, you will select a simple part of water infrastructure鈥攕uch as a straight pipe, a pipe bend, or a small storage tank鈥攁nd use CFD software to simulate how water flows through it. You will compare your computer predictions with known formulas and design guidelines used by engineers. The goal is not only to gain early research experience but also to connect classroom concepts with real-world applications.

By the end of the project, you will:

1. Learn how engineers use computer models to study water systems.
2. Gain hands-on experience running a simple simulation and interpreting results.
3. Communicate your findings through a short report and presentation.
4. Appreciate how research tools like CFD support better design of water infrastructure that we all depend on every day.

This project emphasizes curiosity and creativity over technical complexity鈥攏o prior background in fluid mechanics or programming is required.

By completing this project, students will be able to:

1. Apply fundamental concepts of fluid mechanics (flow regimes, pressure loss, velocity profiles) to practical problems in water infrastructure.
2. Use computational tools (CFD) to set up, run, and interpret simple flow simulations.
3. Compare and validate simulation results against analytical formulas and engineering reference data.
4. Visualize and communicate fluid flow patterns using plots, graphics, and streamlined technical writing.
5. Develop critical thinking by identifying sources of error, uncertainty, and limitations in modeling approaches.
6. Work collaboratively in a research setting, sharing tasks and insights with peers.
7. Experience authentic undergraduate research by posing questions, testing ideas, and making a scholarly contribution through reports or presentations.

Orientation & Planning

• Attend project introduction and overview session.
• Select a case study (pipe, elbow, tank, etc.) independently.
• Define a clear research question and set objectives.
• Conduct initial literature review and identify relevant references.
• Draft a 1鈥�2 page project proposal outlining objectives, expected outcomes, and preliminary plan.

Baseline Calculations & Learning CFD Tools

• Perform hand calculations for simple flow cases (Reynolds number, pressure drop, friction factors).
• Complete CFD tutorials to learn software basics: geometry creation, meshing, boundary conditions, solver setup.
• Run a validation case (e.g., straight pipe) to confirm understanding of the software.
• Document methods and workflow in a personal lab notebook or digital file.

Geometry & Mesh Development

• Build the project-specific geometry in CFD software.
• Develop initial computational meshes and conduct quality checks.
• Run preliminary simulations at low flow rates to verify setup.
• Evaluate results and identify any adjustments needed for the main simulations.

Full Simulation Runs

• Run simulations for all planned flow conditions.
• Refine mesh or solver settings as needed for accuracy.
• Collect simulation data: velocity fields, pressure distributions, head loss, flow patterns.
• Begin analyzing trends and comparing results with theoretical predictions.

Post-Processing & Analysis

• Visualize flow results using plots, contours, and streamlines.
• Compare CFD results with analytical calculations or handbook values.
• Identify sources of error or uncertainty in the simulations.
• Draft figures, tables, and preliminary discussion for the final report.

Report Finalization & Presentation

• Prepare the final technical report including introduction, methods, results, analysis, and conclusions.
• Revise report based on instructor feedback.
• Prepare final presentation materials: poster, slideshow, or short video.
• Present findings to the instructor, peers, or at a departmental showcase.
• Reflect on learning outcomes and personal growth during the project.

Hybrid

Dr. Jie Zhang, jzhang45@kennesaw.edu 

This experimental study explores the dynamic response of a three-story building prototype with and without fluid viscous dampers, with a specific focus on comparing the vibration mitigation performance across a range of different building materials. The primary objective is to enhance structural integrity and stability by identifying the most effective material-and-damper combination for vibration control. The scaled model, with dimensions of approximately 30 cm x 40 cm x 90 cm (1:10 scale), is designed to accommodate and test various structural materials. Initially, the prototype is constructed with steel frames and polymer floors, serving as the baseline for a broader comparative analysis.

To facilitate controlled testing, a custom-designed shake table will be utilized, engineered to produce precise reciprocating motion parallel to the ground while restricting other degrees of freedom. For models tested without dampers, special brackets will provide structural support. The damped models, on the other hand, will be configured as moment-resistant structures to integrate the fluid viscous dampers seamlessly.

Our comprehensive data acquisition system, comprising a waveform generator, an accelerometer, and a vibrometer, will capture detailed vibration data for each material configuration. This will allow for a direct comparative analysis of the materials鈥� inherent vibrational characteristics and the dampers' effectiveness. The newly purchased shake table itself is highly efficient and versatile, capable of testing various model sizes while maintaining a minimal physical footprint.

Ultimately, this research aims to establish a specialized arrangement of dampers and propose a new viscoelastic model that considers the properties of different structural materials. The findings will have significant implications for designing more resilient and safer buildings in seismically active regions, offering valuable insights into the optimal application of vibration mitigation technology across a variety of structural types.

At the end of the project, students should be able to:

• Comprehend the dynamic behavior of various building materials and their response to vibrational forces.
• Develop hands-on skills in the design, construction, and calibration of experimental apparatus, such as a custom shake table.
• Collect, analyze, and interpret complex data from a professional-grade data acquisition system (e.g., accelerometers and vibrometers).
• Articulate how the research findings on fluid viscous dampers and different material properties contribute to the field of civil and structural engineering.
• Define the engineering principles and terminology associated with vibration mitigation and structural integrity.
• Learn how to conduct a comprehensive literature review to understand past research on seismic and vibration control.
• Work effectively as part of a team to solve complex technical challenges.
• Write a research paper to formally document the project's methodology, findings, and conclusions.
• Present their research/creative activity to an audience (e.g., poster, oral presentation, performance, display)
• Reflect on their research project by writing a brief report at the end of their program.

• Study the fundamental principles of structural dynamics and the mechanics of vibration, building on foundational knowledge.
• Conduct a progressive literature review on fluid viscous dampers and their application in civil engineering, presenting findings to the professor during weekly meetings.
• Make observations and collect data from the shake table experiments on the building prototypes, documenting each test meticulously.
• Analyze the collected data and compare it with theoretical models and expected behaviors.
• Become familiar with and practice on software for data processing and analysis, such as MATLAB and Solidworks.
• Prepare an organized weekly report summarizing progress, challenges, and next steps, which will serve as a basis for the final research paper.

Hybrid

Dr. Mohammad Jonaidi, mjonaidi@kennesaw.edu 

Dr. Simin Nasseri, snasser1@kennesaw.edu 

Have you or anyone you know ever been involved in a motor vehicle crash? Have you seen crashes when you travel either as a driver or a passenger? Highway safety is a big concern in Georgia, and the number of crashes (also known as accidents, wrecks, etc.) and associated fatalities and severities have increased considerably in recent years. The economic cost associated with these crashes is in the range of billions of dollars every year. 

Among many causes that lead to these enormous numbers of crashes, driver factors are more common. As we use more and more technology while driving, it has been observed that distracted driving causes significant safety risks to both vehicle occupants and other road users. Despite efforts by the Georgia Department of Transportation (GDOT), and other agencies, distracted-driving-related crashes remain a persistent issue. 

This study analyzes crash data to identify trends, patterns, and contributing factors to distracted driving-related crashes in Georgia. The research will further investigate the relationship between crash occurrence and severity, with the goal of developing recommendations to enhance road safety for all users.

Upon the completion of the project, the student will:

• Understand the basic concepts in conducting research in the broad area of civil engineering
• Understand how a literature review is conducted
• Learn about basic data analysis practices
• Understand how research findings could be derived by analyzing data
• Be familiar with how research could be used to provide guidance on how roadway safety could be improved. 
• Educate themselves and others about the importance of good highway safety practices and the economic benefits associated with reducing crashes

The following are the major tasks the student will engage in, depending on the stage of the project.

• Conducting a literature review about distracted-driving-related safety and understanding the general issues and concerns.
• Gather data related to crashes in Georgia (or part of the state) for the last five years.
• Learn basic statistical methods that could be applied to analyze crash data.
• Apply the most suitable methodology to analyze crash data and identify the critical factors affecting the safety in work zones.
• Write down the activities completed, and record the time spent working on the project
• Write an abstract and participate in the symposium.

Face-to-Face

Dr. Sunanda Dissanayake, sdissan1@kennesaw.edu 

A Software Defined Radio (SDR) is a radio transceiver that is primarily defined in software. It allows radio engineers and researchers to easily control the hardware and implement and configure the physical layer in software. SDRs have been used in many scientific platforms to implement, test, and study different wireless technologies and protocols. In this project, we explore the 5G waveform generation function of the MATLAB 5G Toolbox to generate different uplink-downlink transmission frames.

The 5GNR (5G New Radio) 'knobs' include the bandwidth (BW), the physical resource block (PRB) allocation within the BW, the time slot occupation, the numerology, and the number of transmission layers, and the transmit (Tx) power. The generated waveform is transmitted over the air (OTA) using the GNU Radio software framework in conjunction with Universal Software Radio Peripheral (USRP) or other commercial SDR hardware.

In the proposed work, two SDRs are used to implement the test bed's transmitter and receiver sides. Each SDR is connected by USB 3.0 to a GNU Radio Companion computer. The transmitter and the receiver GNU Radio flowgraphs have a built-in spectrum analyzer (SA) block to visualize the fast Fourier transform of the signal for spectral analysis. We use the 5G Toolbox because of its flexibility to customize the uplink and downlink 5G NR transmission that complies with the 3GPP specifications. In addition, we use another MATLAB function to convert the in-phase and quadrature (IQ) samples to binary points for use with GNU Radio. GNU Radio is a free and open-source software development toolkit that provides the mechanisms and sample signal processing blocks to implement radios in software. It is compatible with low-cost SDR hardware to enable RF transmission.

Furthermore, it has a graphical user interface鈥擥NU Radio Companion鈥攖o assemble radios by connecting blocks. Moreover, we use a Software Defined Radio (SDR) and P21XXCSR-EVB Energy Harvesting module (existing) to conduct experiments on RF energy harvesting from the SDR transmissions. Specifically, we are interested in obtaining insight into joint communication and RF energy harvesting, aiming to transmit as much information as possible under the constraints of providing sufficient RF energy for the receiving device to operate. Experiments that will enable our insights into this problem include comparing communication waveform designs, comparing power emission via USRP, and the effect of environmental conditions like separation distance, presence of nearby in-band transmissions, and other factors. 

There are many high-impact situations where there is loss of telecommunication, for example natural disasters such as hurricanes and places where conventional telecommunication infrastructure is damaged such as in Ukraine, currently. In these cases, software definable, rapid deployable wireless communication systems are urgently needed. Our students will conduct high-impact research to solve crucial telecommunication problems, develop soft and entrepreneurial, and networking skills. We target external grants and industry support. We also prepare the students to use the analytical approach of an engineer, be able to integrate software, hardware and communication systems, and network infrastructures, and develop proof-of-concept projects related to future telecommunications.

The students will develop essential articulation skills as we plan to present the work in seminars and other undergraduate symposiums. 

Specifically, using the proposed test framework described above together with the related tutorials, the participant will be tasked with:

1. Understand the basics of modern communications using the provided tutorials and simulations
2. Develop basic programming and testing of communications using the Matlab framework
3. Perform test simulations using the 5G toolbox in Matlab to learn and utilize the various aspects of the toolbox.
4. Familiarize with and utilize GNU Radio tools
5. Generate a variety of 5G NR frames using the testbed (data generation in Matlab followed by transmission via GNU Radio).
6. Manipulate the test parameters, including system bandwidth, waveform power, system gain, modulation, and subcarrier specifications. Analyze the results
7. Compare the testbed performance to simulation results.
8. Measure the energy harvested via signal transmission.
9. Innovate and Implement strategies to improve energy harvesting while maintaining transmission rates.

Face-to-Face

Dr. Sumit Chakravarty, schakra2@kennesaw.edu 

Cellular communication has advanced significantly over the last four decades, evolving from basic voice services in the first generation (1G) to real-time applications and ultra-fast Internet connectivity in the fifth generation (5G). Accordingly, 5G network provides a significant breakthrough in wireless communication, offering ultra-reliable, ultra-low latency, and massive machine-to-machine communication. As the backbone of emerging technologies such as Internet of Things (IoT), autonomous vehicles, and smart cities, 5G plays a leading role in shaping the future of wireless communication and the world. Accordingly, 5G has emerged with multiple innovative wireless access schemes like massive multiple-input-multiple-output (MIMO) communication, reconfigurable intelligent surfaces (RIS), and non-orthogonal multiple access to cope with the increasing connectivity, reliability, and security demands. Additionally, Artificial Intelligence (AI) will play an important role in enhancing the 5G performance and efficiency. Adopting AI in 5G resource management and optimization allows for a faster dynamic network, which is a dire need for the next-generation cellular networks.

With that said, understanding 5G is a cornerstone to training the next generation of engineers and researchers. This project aims to establish an undergraduate research team in which the students will be engaged in designing, testing, and simulating 5G modules and systems. Additionally, the students will investigate various 5G key-performance metrics, trade-offs, and bottlenecks, as well as brainstorm creative solutions to these issues. The project is designed for first-year students and aims to provide them with a strong technical foundation in wireless 5G communication concepts as well as hands-on exposure to practical wireless communication systems models and simulations. The goal of this project is to promote early engagement in research and development by combining both strong theoretical knowledge of 5G network architecture with practical modeling and simulation techniques. As the project progresses, students will have the chance to learn and apply simulation tools like MATLAB and its 5G-specific toolboxes to model different 5G scenarios. Additionally, they will help simulate advanced 5G systems with advanced resource allocation schemes and compare their simulation results to the current state-of-the-art. 

By the end of this project, the students will have a strong grasp of the 5G fundamentals and technologies. Additionally, they will have an extensive hands-on experience with 5G modules and modeling. The expected student outcomes are summarized as follows.

• Understand the fundamentals of 5G communication systems
• Identify key-performance 5G technologies such as massive MIMO, RIS, and millimeter-Wave (mmWave) communication.
• Comprehend the performance trade-offs and bottlenecks in 5G networks.
• Gain hands-on experience with 5G modeling and simulation tools such as MATLAB.
• Ignite the students鈥� creativity and critical thinking.
• Build the students鈥� teamwork skills.
• Improve the students鈥� communication and presentation skills.
• Gain an early exposure to research work and methodologies.

Students are expected to work on the research project for 5-10 hours per week. This time will be spent on a diverse set of activities depending on the project timeline. In the first weeks of the project, the students will be introduced to cellular communication and generations, 5G, and 5G architecture and technologies. Next, the students will be introduced to modeling. Software programs such as MATLAB will be included, and the students will be able to edit and/or modify existing 5G models. Afterward, the students will be able to model and run specific 5G modules, which will conclude by collecting results and preparing to present at The Symposium of Student Scholars. The list of weekly activities is as follows.

• Attending weekly meetings.
• Preparing weekly progress reports.
• Learning and researching on different 5G fundamentals.
• Working on different 5G simulation tools
• Learning how to write and construct a research paper.
• Preparing presentations for their research findings.

Hybrid

Dr. Radwa Sultan, rsultan7@kennesaw.edu 

This project focuses on developing a prototype that demonstrates how modern techniques can be applied to monitor the health and performance of batteries鈥攖he core of electric vehicles and many renewable energy systems. The prototype will integrate a microcontroller (a small, programmable computer) with a rechargeable battery cell. Students will program the microcontroller to run controlled charge and discharge cycles while monitoring the battery鈥檚 performance. An LCD screen will display the battery鈥檚 State of Charge (SOC), similar to a 鈥渇uel gauge鈥� that indicates how much energy remains in the battery.

What makes this project exciting is the use of an advanced artificial intelligence method, KalmanNet (a neural network鈥揵ased architecture), to estimate the SOC. Unlike traditional methods, which often struggle with noise or sudden fluctuations, KalmanNet combines data-driven learning with neural estimation to provide more accurate and reliable results.

Through this project, students will gain hands-on experience in hardware development, microcontroller programming, energy storage testing, and the application of modern AI algorithms to real-world devices. The outcomes of this work could contribute to safer, more efficient battery management in both electric vehicles and renewable energy applications.

Students participating in this project will develop a well-rounded set of technical, analytical, and research skills that bridge classroom learning with real-world applications.

• Embedded Systems & Programming: Students will learn how to program microcontrollers to run controlled charge/discharge cycles and display key information such as the battery鈥檚 State of Charge (SOC) on an LCD screen.
• Battery Testing & Safety: Hands-on experience with rechargeable battery cells will teach safe handling procedures, performance monitoring, and experimental setup.
• Data Collection & Analysis: Students will practice measuring battery voltage, current, and SOC, then analyze this data to evaluate system performance.
• Artificial Intelligence in Engineering: By working with KalmanNet, students will be introduced to modern AI methods and understand how machine learning can improve real-time estimation compared to traditional techniques.
• Research & Communication: Participants will strengthen their ability to document results, prepare technical reports, and present findings, contributing to knowledge that extends previous international conference work.

These outcomes provide students with practical engineering experience and exposure to cutting-edge methods that are highly relevant to careers in energy storage, control systems, and artificial intelligence applications. By combining hardware prototyping, software development, and research methodology, the project equips students with skills that are valuable for both advanced study and industry practice.

Students working on this project will engage in a variety of weekly tasks that combine research, hands-on experimentation, and technical documentation. These duties are designed to help students build both practical engineering skills and a deeper understanding of energy storage systems.

• Literature & Background Research: Investigate the operation and components of energy storage systems, with a focus on current industrial standards. Students will also use library resources to explore existing research relevant to their work.
• Modeling & Simulation: Learn to use engineering software tools such as MATLAB/Simulink to model, analyze, and simulate electrical systems under different operating conditions.
• Experimental Work: Conduct laboratory experiments, operate test equipment, and collect performance data from prototype systems. Students will also participate in testing and troubleshooting experimental prototypes.
• Technical Writing & Communication: Prepare high-quality weekly reports that summarize research findings, simulation results, and experimental progress. These reports will help develop professional writing and communication skills expected in academic and industry settings.

By completing these weekly responsibilities, students will gain structured experience in energy storage research, modeling, and experimental validation. The balance of independent learning, guided mentorship, and team collaboration ensures steady progress while building a strong foundation for future research and engineering practice.

Hybrid

Dr. Yousef Mahmoud, ymahmoud@kennesaw.edu 

Additive manufacturing, also known as 3D printing, is transforming how metal parts are manufactured for industries like aerospace, energy, and healthcare. Two standard methods, Laser Powder Bed Fusion (LPBF) and Bound Metal Deposition (BMD), build parts layer by layer, making it possible to create complex shapes that would be difficult or impossible with traditional manufacturing. While these techniques can produce high-performance components, the finished surfaces contain internal defects, rough areas, residual stress, or uneven material structures. These imperfections affect the product's strength, making the part more vulnerable to corrosion and shortening its lifespan.

This project focuses on improving the surface quality and overall performance of 3D-printed metal parts using three advanced finishing methods. TIG remelting uses a controlled welding arc to melt just the surface, sealing pores, improving the microstructure behavior on the surface level, and smoothing out the finished product. Electroplating adds a thin, even metal layer to improve corrosion resistance and durability. Sol鈥揼el coating applies a ceramic-like layer that provides extra protection against heat, chemicals, and environmental damage. The research will study each of these treatments individually and in combination, measuring their effects on corrosion resistance, surface smoothness, impact strength, tensile strength, hardness, and performance under steam cycle conditions. The project will also look at how these processes affect the microscopic structure of the metal (microstructure behavior) and how they influence annealing behavior, which is the way the material changes when heated and cooled to improve its properties. A leak test will be performed to assess whether the treatments effectively seal the part against fluid or gas leakage.

By comparing the results, the project aims to identify the most effective single treatment or combination of treatments for producing stronger, longer-lasting 3D-printed metal parts. The findings could lead to more reliable components for demanding real-world uses ranging from aircraft engines to power plant equipment, helping these parts last longer, perform better, and withstand harsher conditions.

Students will participate in laboratory work to develop research and testing techniques based on existing standards, while also creating new testing methods due to limited prior research in this area. They will learn how to effectively present their ideas and research to both technical and non-technical audiences. The skills acquired from this program will benefit those aspiring to become researchers or pursue graduate studies. 

Students will carry out academic research and perform physical experiments. They are also expected to participate in meetings and contribute significantly as team members. 

Hybrid

Dr. Aaron Adams, aadam224@kennesaw.edu 

Patient鈥搉urse interactions are central to high-quality healthcare delivery, directly influencing patient satisfaction, knowledge retention, self-care management, and emotional well-being. As health systems increasingly implement virtual care models such as the Virtual Nursing Care Model (VNCM), it is critical to examine how patients engage with and respond to both bedside and virtual nurse education. Prior research has generally focused on single aspects of these interactions such as verbal communication or visual attention, without exploring their interplay in a multimodal context. This study addresses this gap by using a multimodal approach, integrating eye-tracking and prosodic vocal analysis, to comprehensively assess patient engagement and emotional response during education sessions with bedside and virtual nurses. By examining these factors together, this research seeks to provide a holistic understanding of how patients engage with their caregivers and how this engagement differs between bedside and virtual settings. The insights gained will inform the design of more effective patient education strategies in virtual healthcare settings to mimic in person interactions.

Through this project, students will gain hands-on experience in the ABCs of research; from learning how studies are designed to collecting and analyzing data. Students will:

• Learn what research is and how to ask meaningful questions.
• Complete CITI certification and understand the role of the IRB in protecting participants.
• Learn how to conduct a literature review to understand what is already known and identify research gaps.
• Work with advanced tools like eye-tracking and voice analysis software.
• Develop skills in data analysis, teamwork, and communication by going into the hospital, interacting with patients, and collecting real-world data.

By the end, students will have a strong foundation in research methods, communication, and problem-solving that can be applied to future opportunities in healthcare, engineering, and beyond.

Students will participate in a variety of hands-on research activities each week. While their responsibilities may differ from week to week, they will be involved in activities such as:

• Attend and actively participate in weekly meetings by sharing updates, asking questions, and helping plan next steps with the research team.
• Research skill building by learning how to conduct literature reviews, read and summarize journal articles, and discuss findings with peers and mentors.
• Training and certification through completing CITI training and learning about IRB requirements for working with human subjects.
• Study preparation by helping set up equipment (eye tracking, voice analysis tools) and practicing data collection techniques.
• Hospital visits to observe and assist with data collection, including interacting with patients during their education sessions.
• Data collection and coding by recording observations, gathering data, and beginning to code patient interactions for analysis.
• Data analysis and reflection to learn how to interpret results, reflect on findings, and connect them back to the research questions.
• Poster and presentation preparation by working with the team to design and present a research poster for the Undergraduate Student Research Symposium.

Face-to-Face

Dr. Awatef Ergai, aergai@kennesaw.edu 

Everyone loves to have a well-manicured green lawn but it is hard to manage weeds. This endeavor is all about smarter, simpler, and more efficient lawn care using artificial intelligence (AI).

Weeds are foreign plants that have the potential to damage your lawn's appearance and health. It can prevent overgrowth, reduce the amount of poisonous chemicals used, and save time. But identifying weeds鈥攅specially when they blend in with grass鈥攊s challenging even for experts. Technology is where this dilemma is addressed.

In this project, we are developing a system that uses computer vision, an area of AI that allows computers to "see" and understand pictures, to recognize weeds in a lawn automatically. We will be particularly using a powerful AI utility called Detectron2, which can recognize and outline the edges of objects (like weeds) in pictures with great accuracy. By training this tool to recognize hundreds of examples of common lawn weeds, we aim to create an efficient and effective detection system.

Our long-term goal is to use this AI model in real-world applications鈥攍ike hooking it up to a smartphone app or even a robot that would be able to apply the chemicals itself. This would allow homeowners and landscapers to more effectively take care of their lawns with less effort and fewer chemicals. 

The students will assist in gathering and annotating pictures of some common lawn weeds, learn about how AI systems are trained to recognize patterns in images,assist in testing and perfecting the weed detection system,get practical experience in one of the most rapidly developing fields of science and technology.

Students don't need any programming experience and will be taught what they will have to learn. If they are fascinated by AI, such as working in the field or with images, and would like to help develop a new and emerging technology that could make daily life easier, this project is a great opportunity.

Students who take part in this project will gain hands-on experience in one of the most exciting and fastest-growing fields of our era鈥攃omputer vision and artificial intelligence (AI). As students go through the process of developing the weed detection system, they will gain a set of practical skills that are both highly relevant to research and in industry demand.

The main outcomes are:

Image Data Collection & Annotation: Students will learn to gather and annotate images of different lawn weeds, a prerequisite for creating accurate AI models.

Introduction to Artificial Intelligence and Machine Learning: Students will gain an introductory understanding of how machines learn to recognize patterns in data, especially through visual inputs like pictures.

Hands-on with Detectron2: Students will be introduced to Detectron2, the latest state-of-the-art deep learning framework developed by Facebook AI. They will assist in training, testing, and fine-tuning object detection models.

Problem Solving & Critical Thinking: By going through the discussion of real problems such as weed detection in mixed lighting and grass environments, students will enhance their creativity of thought and problem-solving.

Collaboration and Research Communication: Students will work closely with a faculty mentor and peers, learning how to present results, give feedback, and strive toward a common goal of research.

Ethics and Environmental Sustainability in AI: Students will engage in open discourse about the broader impact of AI use on the environment, data privacy, and environmentally friendly lawn care practices.

By the end of the project, students will have been introduced to the world of research and artificial intelligence, gaining skills that can be transferred to potential future college or career prospects.

Students working on this project will have a mix of hands-on, collaborative, and learning activities each week.

Sample weekly activities could be:

Image Collection and Field Work (1鈥�2 hours/week)
Students will take photographs of various types of weeds in lawn environments, either at designated locations or using smartphones from secure, accessible positions. This helps in the aggregation of the image data set to be used for training the AI model.

Image Annotation & Dataset Preparation (1-2 hours/week)
Using easy-to-learn labeling tools, students will annotate and categorize weeds in images. This activity demonstrates the importance of quality data in training AI systems.

Workshops & Mentoring Sessions (1 hour/week)
Students will participate in guided sessions in order to learn computer vision, artificial intelligence, and how Detectron2 works. They also provide time for questioning and feedback on the research process.

Team Collaboration (1鈥�2 hours/week)
Students will meet with the faculty mentor and team members to discuss progress, fix bugs, and plan next steps. They can also provide short updates or share new insights.

Model Testing & Feedback (1 hour/week)
During development of the AI model, students will aid in testing the accuracy of how it detects weeds in new images and offer feedback on how to enhance. This exposes them to measuring model performance and basic problem-solving in AI use.

These weekly activities are designed to be challenging and accessible to first-year students with no prior research background.

Hybrid

Dr. Sathish Kumar Gurupatham, sgurupat@kennesaw.edu

The project focuses on investigating the aerodynamic behavior of tandem wing configurations, a topic of significant importance in the design of advanced aircraft and unmanned aerial vehicles (UAVs). The primary objective is to understand the complex interactions between the vortices generated by the leading wing and their impact on the aerodynamic performance of the trailing wing. This research is driven by the need to optimize tandem wing arrangements, which can potentially offer higher lift-to-drag ratios, improved stability, and enhanced maneuverability in various flight regimes.

The study will employ advanced computational fluid dynamics (CFD) techniques to simulate the flow fields around the tandem wing configuration. The numerical analysis will be conducted using high-fidelity CFD software, with turbulence modeling approaches such as Reynolds-Averaged Navier-Stokes (RANS), Large Eddy Simulation (LES), or Detached Eddy Simulation (DES) to capture the intricate details of vortex formation, shedding, and interaction. The simulation results will provide detailed insights into the wake dynamics, vortex trajectories, and their influence on the pressure distribution, lift, drag, and overall aerodynamic efficiency of the tandem wing arrangement.

A key aspect of the project is the parametric study, where different wing configurations, spacing, and angles of attack will be analyzed to determine their effect on vortex interaction and aerodynamic performance. The study will also explore the effects of varying Reynolds numbers to understand the performance characteristics across different flight conditions. The results from the numerical simulations will be validated against available experimental data or empirical correlations to ensure accuracy and reliability.

The outcomes of this research are expected to contribute to the design guidelines for tandem-wing aircraft, enabling the development of more efficient and high-performance aerial vehicles. The findings will have implications for both military and civilian applications, where tandem wing configurations are being considered for their potential advantages in specific mission profiles.

Overall, this project aims to advance the understanding of vortex interactions in tandem wing arrangements, providing valuable insights that can lead to innovative design strategies in the aerospace industry. The research will culminate in the development of comprehensive design recommendations and potential optimization strategies for future tandem-wing aircraft designs.

Students participating in this project will develop a robust set of skills and techniques that are essential for careers in aerospace engineering and related fields.

Firstly, students will gain hands-on experience with advanced computational fluid dynamics (CFD) software, learning how to set up, run, and analyze simulations of complex aerodynamic systems. They will become proficient in using CFD tools such as ANSYS Fluent, which are industry standards, and will learn how to apply different turbulence modeling approaches like Reynolds-Average Navier-Stokes Equations (RANS), Large Eddy Simulation (LES), or Detached Eddy Simulation (DES) to capture the intricate details of vortex dynamics.

In addition to technical software skills, students will develop strong analytical and problem-solving abilities. They will learn to interpret the complex data generated from simulations, extracting meaningful insights into the behavior of vortices in tandem wing configurations. This process will involve understanding and applying fundamental aerodynamic principles, as well as critically evaluating the accuracy and reliability of their results.

Moreover, students will enhance their research and experimental design skills by conducting a parametric study to explore the effects of different wing configurations, spacing, and angles of attack. They will learn how to design experiments that test specific hypotheses, control variables effectively, and draw conclusions based on empirical data.

Collaboration and communication skills will also be a significant focus. Students will work in teams, learning how to collaborate effectively, share responsibilities, and integrate their contributions into a cohesive research effort. They will also develop the ability to communicate their findings clearly and concisely, whether through written reports, presentations, or discussions, which is critical for any professional or academic career.

Overall, this project provides students with a comprehensive skill set that prepares them for future challenges in both academic and industry settings.

Throughout the two-semester duration of this project, students will engage in structured activities designed to build their skills, knowledge, and research capabilities progressively. The timeline is considering the program starts the week of September 29th:

Semester 1

Weeks 1鈥�2: Orientation and Background Research

• Introduction to fundamental concepts in aerodynamics and computational fluid dynamics (CFD).
• Targeted literature reviews on tandem wing configurations and vortex interactions.

Weeks 3鈥�4: CFD Software Training

• Focused training sessions with ANSYS Fluent (or equivalent).
• Completion of essential tutorials covering geometry setup, meshing, and baseline simulation runs.

Weeks 5鈥�8: Simulation Setup and Initial Analyses

• Setup of initial tandem wing simulations with simplified configurations.
• Execution of preliminary runs to study vortex formation and flow field characteristics.
• Ongoing check-ins to ensure technical progress and address challenges.

Week 9 (Final Week of Semester): Interim Review

• Students present preliminary results and lessons learned.
• Feedback session to prepare for more advanced work in Semester 2.

Semester 2

Weeks 1鈥�4: Advanced Simulation Runs

• Refinement of models and extension to more complex wing configurations.
• Parametric studies examining spacing, angles of attack, and Reynolds number effects.

Weeks 5鈥�8: Validation and Comparison

• Validation of simulation results against experimental data or empirical models.
• Comparative analyses to assess accuracy and reliability.

Weeks 9鈥�12: Documentation and Presentation Preparation

• Compilation of results into a comprehensive technical report.
• Development of presentations for internal reviews, conferences, or publication submissions.

Weeks 13鈥�15: Final Review and Reflection

• Formal presentation of research findings.
• Reflection on the research process and discussion of potential future directions.

Hybrid

Dr. Gaurav Sharma, gsharma3@kennesaw.edu 

How can a short medical video help doctors make quicker, safer decisions? In this First-Year Scholars project at KSU, you鈥檒l help transform real endoscopy videos into clear, useful information using approachable tools from machine learning. We partner with clinicians and researchers at Emory University to explore questions like: How wide is the esophagus opening? How does it change over time in a procedure? Can simple visuals help physicians spot issues sooner?

No prior coding or medical background is required鈥攋ust curiosity and a willingness to learn. We start with guided, hands-on mini-lessons in Python and basic computer vision. You鈥檒l learn how to (1) organize video frames, (2) apply beginner-friendly ML methods to highlight the center 鈥渙pening鈥� in each image, and (3) create easy-to-read overlays and graphs that turn pixels into understandable measurements. Along the way, we鈥檒l cover research basics (project notebooks, version control), teamwork, and the ethics of working with de-identified medical data.

What you鈥檒l do: 1 Work with real, de-identified endoscopy videos provided through our Emory collaboration. 2 Help label or verify images, run analysis notebooks, and check model outputs for quality. 3 Convert results into simple visuals鈥攖hink short demo videos, before/after overlays, and charts (e.g., violin plots) that summarize measurements across frames. 4 Share your findings in clear, everyday language for clinicians and peers.

What you鈥檒l gain: 1 Practical experience with AI/ML that you can showcase in a portfolio, poster, or presentation. 2 Confidence using Python for real-world health applications. 3 Mentorship from KSU faculty and exposure to clinical perspectives from Emory collaborators. 4 A meaningful project that connects computing to people鈥檚 health.

If you鈥檙e excited by the idea of turning raw video into insights that could support better care, this project is for you. Join us to learn by doing, contribute to an active KSU鈥揈mory research collaboration, and help make medical imaging more informative and accessible.

Students will leave this project with practical, r茅sum茅-ready skills in computing, research, and communication. You will learn Python from the ground up using Jupyter notebooks and apply it to real images and video. Along the way you鈥檒l practice core computer-vision techniques (thresholding, morphology, contour analysis) and basic machine-learning workflows for measurement and quality control. You鈥檒l handle data responsibly鈥攚orking with de-identified clinical videos, keeping clear records, and writing reproducible code with Git/GitHub. You鈥檒l also learn how to turn results into insight: generating overlays and short demo videos, building summary figures (e.g., violin plots of diameters), and checking your work with error analysis and simple statistics.

Because this is research, you鈥檒l develop habits that courses rarely teach: framing a testable question, comparing alternative methods, tuning parameters thoughtfully, and documenting why decisions were made. You鈥檒l practice debugging and troubleshooting, design small experiments to validate your approach, and interpret trade-offs between speed, accuracy, and robustness. Regular mentorship helps you learn to give and receive technical feedback, plan sprints, and iterate efficiently.

Communication matters just as much as code. You鈥檒l learn to explain methods and findings in clear, non-jargon language for clinicians and peers, and to present your work through a brief report, a poster, or a short talk. Collaboration with KSU mentors and our Emory partners will build your confidence working across disciplines and with real stakeholders. By the end, you will have an end-to-end portfolio piece鈥攁 documented analysis pipeline, example outputs, and concise visuals鈥攖hat demonstrates beginner-friendly AI applied to a real medical problem.

Team check-in. We start with a short stand-up to set goals, review progress, and assign tasks from our shared checklist (data prep, coding, figure updates, writing).

Mini-lesson + hands-on lab. You鈥檒l get a bite-sized tutorial (e.g., Python basics, image thresholding, simple machine-learning checks, or version control with Git/GitHub) and immediately apply it to real endoscopy frames.

Quality checks. Compare model output to visual expectations: Is the green lumen region correct? Does the diameter label make sense? You鈥檒l flag tricky cases, adjust parameters, and re-run as needed.

Annotation or labeling. When helpful, you鈥檒l label a few images to improve evaluation or provide examples for training/validation.

Results and reflection. Update your notebook with what you tried, what worked, and what to try next. Create quick visuals (before/after overlays, violin plots of diameters) and add them to the shared folder.

Mentor feedback (10鈥�30 min). Get asynchronous comments on your pull request or meet briefly with a mentor to unblock issues. About every other week we鈥檒l sync with our Emory collaborators to discuss clinical relevance and prioritize next steps.

Professional habits (sprinkled throughout). You鈥檒l practice ethical data handling, clear documentation, and collaborative workflows (issues, branches, pull requests). Toward the end of the term, time shifts to polishing a short poster or demo video for sharing your results.

Total weekly time is designed to be beginner-friendly and flexible, with most work completed in short, focused blocks you can schedule around classes.

Hybrid

Dr. Lei Shi, lshi@kennesaw.edu 

Insect locomotion is a unique model for the study of how nervous systems produce locomotor patterns and how we can build bio-inspired robots. Drosophila (fruit flies) have complex and adaptive locomotor behaviors despite their small size, which make them a great system to study. Their compact nervous system, combined with well-developed experimental methods for genetic manipulation and neurosurgery, allows researchers to directly connect neural changes with measurable behavioral outcomes. Performing brain surgeries and then analyzing the changes in walking would allow us to understand how the brain controls walking and will inform us how we can build better adaptive robots.

The goal of this project is to analyze how brain surgeries in fruit flies affect locomotion patterns, and to compare these patterns between tethered and untethered conditions, while also examining variations across sex and age groups. During this experience, students will learn how to perform microsurgery on a fly under supervision, which will be followed by a behavioral assay later on. After the surgeries are completed, students will track the gait patterns in a locomotion compensator system, which will be used to record high-resolution walking videos after surgery. Additionally, the students will run a side-by-side comparison of tethered versus non-tethered, male versus female flies to see how locomotion strategies change between these experimental conditions. This will allow us to understand the constraints as well as the adaptive locomotion of the flies post-surgery.

The primary output of this project an analysis of the different gait parameters, such as stride frequency, step symmetry, stability, and turning that are altered after surgery, and the side-by-side comparison with a tethered condition will also help elucidate how different constraints (such as being tethered) affect locomotor strategies. This will give context for any post-surgery recovery observed in flies. As such, this project will generate original data that is important for understanding insect neurobiology and for bio-inspired robotics, where stability and adaptability are design considerations.

This project also provides training opportunities for undergraduate students to get hands-on experience with experimental design, microsurgery procedures, and quantitative data analysis. During the experience, the students will be able to contribute to an existing line of research while also learning interdisciplinary research skills in both biology and engineering, to help prepare them for more advanced research work in the future. The results of this project will also be important for the neuroscience and robotics communities.

The project aims to equip students with essential skills for both academic advancement and professional success. By engaging in hands-on research, teamwork, and guided mentorship, they will gain experience in the following areas:

• Learn basic fruit fly microsurgery techniques under direct mentorship.
• Gain experience in behavioral data collection using video-based tracking and locomotion compensators.
• Acquire skills in data analysis and quantitative gait metrics (step frequency, stride length, symmetry).
• Experience in modifying and operating a locomotion compensator, including mechanical and electronic components.
• Understand the impact of experimental design variables (tethering vs. non-tethering) on locomotor behavior.
• Enhancement of teamwork and collaboration skills through engagement with peers and mentors.
• Familiarity with software tools for simulation, analysis, and data visualization.
• Develop scientific problem-solving, troubleshooting, and hypothesis-testing abilities.
• Strengthen communication skills through documentation, weekly presentations, and final reporting.
• Engage in interdisciplinary work connecting engineering, biology, and neuroscience.

Over the course of the project, students will take part in a range of activities that gradually build their skills while advancing the overall research goals. Their weekly responsibilities will include a mix of hands-on experiments, data analysis, team discussions, and careful documentation, providing a well-rounded research experience. Examples of these activities include:

• Research Activities:
  • Microsurgery Training: Students will learn and practice fruit fly brain surgery on specimens under supervision, gradually developing skills.
  • Sample Preparation and Conducting Experiment: Operate the transparent omnidirectional locomotion compensator (TOLC) to record the gait of both tethered and freely moving flies, before and after surgery.
  • Data Analysis: Use software tools such as MATLAB and Python to analyze data and extract locomotion metrics, including speed, gait symmetry, and turning behavior.
  • Comparative Experiments: Run controlled trials with tethered and non-tethered flies to compare behavioral outcomes.
• Weekly Meetings and Documentation:
  • Weekly Progress Meetings: Students will participate in weekly team meetings with their peers and mentors to discuss progress, troubleshoot challenges, and plan for the upcoming week.
  • Research Log Maintenance: Students will maintain a detailed research log, documenting their weekly activities, observations, and any modifications made to the experimental setup.
  • Report Writing: As the project progresses, students will draft sections of the final research report, summarize their findings, and reflect on their experiences. These duties will provide students with a comprehensive understanding of the research process, from data collection to analysis and reporting, ensuring they are fully engaged in each aspect of the project.

Face-to-Face

Dr. Dal Hyung Kim, dkim97@kennesaw.edu 

What if robots don't just move with motors and gears, but with muscles inspired by biology made of smart materials? This project explores the creation of such an intelligent soft robotic actuator, a new type of "muscle" for future robots. Instead of relying on rigid parts, these actuators are built from shape memory alloys (SMA), materials that can 鈥渞emember鈥� their original shape and return to it when heated or cooled. When combined with clever designs, SMA-based actuators can bend and flex in ways that mimic biological muscles. Even more exciting feature of such robot is - these robotic muscles can do more than just move. They can also process information using artificial intelligence, giving them a kind of built-in intelligence.

The actuator is built using SMA, which has a unique ability to 鈥渞emember鈥� its original shape and recover it when triggered by heat. When paired with other materials in a special design, the SMA can bend and move like biological muscles. Using this property, it can also be designed to 鈥渢hink鈥� in a simple way by relying on its natural dynamics and movement for computation. This approach is part of a comparatively new branch of artificial intelligence known as physical reservoir computing. Instead of relying solely on software-based algorithms, this process allows the physical system itself to process information. In this case, the bending and vibrations of the actuator can be used to solve signal classification problems, such as distinguishing patterns similar to a cat versus dog recognition task, as well as prediction of its own movement or similar predictive machine learning problems, or other pattern recognition tasks.

Students will explore the actuator through two complementary paths. The first is simulation and modeling, where they will use computer tools to predict how the actuator behaves under different conditions. The second is hands-on experimentation, where they will build prototypes and test their performance in the lab. This dual approach allows students with different interests to contribute, where some may focus on design, simulation, and data analysis, while others may focus on fabrication and testing.

These intelligent muscles could transform the future of robotics. Potential applications include bioinspired robotic limbs that move more like natural muscles, medical and assistive devices that adapt to patients鈥� needs, wearable exoskeletons for rehabilitation, and even soft robotic explorers that can move through environments too dangerous or complex for rigid machines.

Students who join this project will gain both technical and professional skills that prepare them for future opportunities in research, engineering, and innovation. From the very beginning, they will learn how to approach open-ended problems where the answers are not already known. This experience will help them grow into independent thinkers who can contribute ideas as well as follow procedures.

On the technical side, students will be introduced to the fundamentals of soft robotics and the design of robotic muscles. Some students may focus on CAD design, computer simulation and modeling, where they will learn how to design, and predict the behavior of smart materials using software tools. Others may focus on hands-on experimentation, working directly with shape memory alloys, assembling prototypes, and running performance tests in the lab. Along the way, all students will practice collecting and analyzing data, comparing simulation predictions with real-world results, and learning how robotic systems can be used for both motion and simple information processing. At the same time, students will build professional skills that are essential for success in any field. They will learn how to work collaboratively in a research team, how to solve problems creatively when unexpected challenges arise, and how to communicate their work to both technical and non-technical audiences. They will also practice explaining the broader significance of their results, connecting their contributions to future applications such as robotic prosthetics, assistive exoskeletons, and soft robotic explorers.

By the end of the project, students will not only have new skills but also the confidence of knowing that their work contributed to an active area of research. This experience will serve as a foundation for future academic projects, internships, or careers in science and engineering.

Students will actively contribute to both simulation and experimentation, with weekly activities tailored to their interests and strengths.

Simulation Track (for students interested in simulation/data):

• Preparing CAD design.
• Running simple dynamics simulations of the actuator鈥檚 bending and vibration.
• Analyzing how different parameters (geometry, input signals, or temperature) affect performance.
• Comparing simulation predictions with experimental data.

Experimental Track (for students interested in hands-on lab work):

• Building prototype actuators using SMA wires.
• Running lab tests by applying electrical signals using Arduino and simple electronic components.
• Recording actuator movement with cameras and sensors.

Shared Activities:

• Brainstorming design ideas and applications.
• Documenting progress with notes, sketches, and photos.
• Collaborating as a team to connect simulation and experimental findings.
• Preparing presentation materials that highlight both modeling and real-world demonstrations.
• Discuss their findings and ideas to the PI.

By the semester鈥檚 end, students will present a combined story of simulations guided experiments, and how experiments validated or challenged simulations demonstrating a complete cycle of research that could inform the design of next-generation robotic muscles.

Hybrid

Dr. Md Raf E Ul Shougat, mshougat@kennesaw.edu 

What we do 
We build real鈥憈ime AI and robotics systems that run on the edge. Projects include AGICore (LLM鈥慻uided real鈥憈ime control), ARTISAN (intelligent space awareness with digital twins), Healing-Hands (emergency room companion), assistive/rehabilitation devices, biomedical devices and autonomous systems.

Who we鈥檙e looking for
Students who are dedicated to research, 蝉别濒蹿鈥憁辞迟颈惫补迟别诲, and eager to learn and grow with the lab. Curiosity, consistency, and follow鈥憈hrough matter more than prior experience.

What you鈥檒l do (starter tasks)

• Bring up sensors (camera, mic, IMU) and log small datasets.
• Run baseline models and record results (accuracy, latency).
• Write clean, commented Python/C++ and use Git properly.
• Test on a robot (quadruped or arm) or in simulation; file concise reports.
• Contribute figures or a short demo for a poster or internal showcase.

What you鈥檒l learn

• Agentic AI (sense, device, act and interact)
• Practical AI/robotics workflows: data 鈫� model 鈫� deployment.
• 搁翱厂鈥�2 basics, multi鈥憇ensor calibration, and simple sensor fusion.
• Reproducible research habits: branches, issues, experiment logs.
• Biomedical devices from the ground up. 
• Communicating results with clear visuals and short summaries.

Expectations (non鈥憂egotiable)

• Attendance is required at weekly stand鈥憉ps and check鈥慽ns.
• Contribution is required every week (code, data, analysis, or documentation).
• Reply to lab messages promptly and close tasks you own.
• Follow lab templates for READMEs, PRs, and experiment notes.
• Document and contribute to Microsoft Team for the group. 

Support you鈥檒l get

• Onboarding guides, starter repos, and small 鈥渇irst issues.鈥�
• Regular code reviews and quick feedback loops.
• Pairing with a mentor (grad student or senior scholar).

How success looks
By the end of the term, you will have (1) a merged contribution to an active project, (2) a small public artifact (poster/demo/short paper section), and (3) the confidence to take on a bigger subsystem next semester.

If this sounds like you鈥攁nd you鈥檙e ready to show up, contribute, and grow鈥攚e鈥檇 love to work with you.

We are also part of Prototype Living of Tomorrow (PLOT) Initiative at 麻豆传媒社区 University, a newly formed innovation hub.

You will learn to:

• Move from data 鈫� model 鈫� deployment on edge hardware.
• Use Python/C++, Linux, Git/GitHub, and basic containers.
• Bring up and calibrate cameras, microphones, and IMUs; time鈥憇ync and fuse sensor streams.
• Train/evaluate models; track metrics (precision/recall, latency, power).
• Implement 搁翱厂鈥�2 nodes, behavior trees, and task pipelines.
• Map LLM/vision鈥慙LM intents to deterministic skills for real鈥憈ime control.
• Build simple digital鈥憈win dashboards for telemetry and playback.
• Communicate results with clear figures, short write鈥憉ps, and a poster/demo.

Many tracks to pick from (you can change anytime):

1. Perception (detection, tracking, model optimization).
2. Planning & Control (navigation, behaviors, evaluation).
3. Systems & Edge (pipelines, profiling, deployment).
4. Human鈥慍entered Tools (dashboards, labeling, evaluation UX).

What you鈥檒l take away:

• A merged contribution to an active AGICore/ARTISAN repo.
• A public artifact (poster/demo/short paper section).
• Confidence using professional tools and collaborating across disciplines (biomedical, cybersecurity, networking, infrastructure) .

Every week you will:

• Attend the required stand鈥憉p (status, blockers, next steps).
• Meet briefly with a mentor for targeted feedback.
• Work from a scoped issue: implement a feature or run a focused experiment.
• Collect or label data; keep tidy folders documentation and metadata.
• Run tests on a robot (quadruped/arm) or in simulation; record latency/accuracy.
• Push code to a branch, open a pull request, and respond to review.
• Develop hardware, 3D print, model and create systems. 
• Log results (what you tried, metrics, screenshots/plots) and draft a short summary.
• Update your task board/Teams group and close issues you own.

Expectations:

• Attendance is required; contribution is required every week (code, data, analysis, or documentation).
• Communicate promptly on lab channels; ask for help early.
• Use lab templates for READMEs, experiment logs, and figures.

We are looking for students who are dedicated to research, 蝉别濒蹿鈥憁辞迟颈惫补迟别诲, and eager to learn and grow with the lab. Attendance and weekly contribution are required.

Hybrid

Dr. Razvan Voicu, rvoicu@kennesaw.edu 

Dr. Muhammad Tanveer, mtanveer@kennesaw.edu 

The loss of a hand can greatly affect a person鈥檚 daily life, making even simple tasks like picking up objects, holding a cup, or writing very challenging. Traditional prosthetic hands often look rigid and lack the flexibility and adaptability of a natural human hand. Our project aims to change this by creating a new kind of prosthetic hand that is soft, intelligent, and more responsive to real-world needs.

This project focuses on designing and developing a soft robotic prosthetic hand that uses smart technology to function more like a human hand. Unlike rigid prosthetics, soft robotics uses flexible materials that can bend and adapt, giving the prosthetic a more natural range of motion. This makes it not only more comfortable to wear but also more effective at handling delicate or irregularly shaped objects.

To achieve this, we will begin by creating a simple computer-based model of the hand. This model allows us to test different designs and movements virtually before building the physical version. By using this numerical approach, we can improve the efficiency of the design process and predict how the hand will perform in real life.

Another important feature of our prosthetic hand is its integration with computer vision. By using a small camera, the hand will be able to 鈥渟ee鈥� objects in front of it. With the help of intelligent algorithms, the system will identify objects and decide the best way to grasp them. For example, it will recognize the difference between a pen, a bottle, or a fragile object like an egg, and adjust its grip accordingly. This adds a level of intelligence and adaptability that most prosthetic hands do not have.

Once the design and computer model are ready, we will use 3D printing to build the actual prosthetic hand. 3D printing allows us to create custom parts that are lightweight, cost-effective, and tailored to the design. We will then integrate the hardware, including the camera and electronic components, to bring the prosthetic hand to life.

In the end, our goal is to develop a prosthetic hand that is not only functional but also affordable and accessible to more people. This project represents an important step forward in combining soft robotics, artificial intelligence, and 3D printing to create smarter, more human-like prosthetics that can truly improve lives.

• Prototyping Skills 鈥� Build real-world experience with 3D modeling, 3D printing, and hardware integration.
• AI & Computer Vision 鈥� Apply artificial intelligence to make the prosthetic hand recognize and interact with objects.
• Interdisciplinary Learning 鈥� Combine robotics, biomedical engineering, and computer science to solve complex problems.
• Research Experience 鈥� Contribute to innovative work with opportunities for conference presentations or publications.
• Human-Centered Impact 鈥� Design technology that improves daily life for people with limb loss.

Each week, students will actively participate in a blend of design, research, and hands-on experimentation that balances technical learning with creativity. Early weeks will focus on concept development and digital modeling, where students will sketch designs, explore soft robotic mechanisms, and create simple computer-based models of the prosthetic hand. This stage will also involve reviewing background literature and discussing design ideas with faculty mentors and peers.

As the project progresses, weekly duties will shift toward simulation and coding tasks, where students will test hand movements virtually and begin integrating computer vision algorithms. This will provide direct experience with programming tools and allow students to see how cameras and software can help the hand 鈥渞ecognize鈥� and interact with different objects.

Students will also engage in prototyping and fabrication, using 3D printers to bring their designs to life. Weekly sessions will include building, assembling, and testing components, offering practical engineering experience. Students will refine designs through an iterative process, making adjustments based on performance results.

Another key duty will be collaborative problem-solving. Weekly meetings will encourage brainstorming and knowledge-sharing across disciplines, ensuring that every student contributes to the integration of hardware, software, and design. Students will also document progress through lab notes, short reports, or presentations, gaining practice in communicating technical work clearly.

By the later weeks, duties will focus on system integration and testing, where students will combine the physical hand with the camera and electronic systems, evaluate performance, and troubleshoot challenges. Each week builds on the last, ensuring that students not only learn new technical skills but also see the tangible progress of creating a functional, intelligent prosthetic hand.

Face-to-Face

Dr. Turaj Ashuri, tashuri@kennesaw.edu 

Dr. Amir Ali Amiri Moghadam, aamirimo@kennesaw.edu