Project Description:

The problem of aerial object detection can be applied to wide variety of problems relevant to GTRI sponsors such as Automated Target Recognition, target detection, and tracking problems. Neural Architecture search through genetic algorithms is a viable approach to model optimization for many tasks. The surrogate evaluator will be added to this pipeline and will be trained to predict performance metrics formodel architectures, and be used as a preliminary evaluation technique. This approach applied to neural architecture search should greatly improve optimization compute time by reducing the number of expensive training operations required to find algorithms that perform well at object detection. Furthermore, we can use back-propagation to generate new algorithms for the optimization search, which is a new approach that has potential to create targeted algorithms.

Citizenship Required:

US-Persons Only

In-Office Time Required: 100%

Competencies:

Computer Science, Machine Learning, Deep Learning, Genetic Programming

Project Description:

Acoustic IDentification (AID) tags are engineered structures that are deployed at known positions underwater to mark points of interest, or provide navigational assistance to autonomous underwater vehicles (AUVs). They work by backscattering a unique acoustic signature based on their design, when insonified using SOund NAvigation and Ranging (SONAR) hardware. The current design iteration of an AID tag comprises of a set of nested hemispherical shells, flooded with water in the space between the shells. This design produces backscatter similar to a barcode that can be uniquely detected using signal processing and machine learning techniques.

Over the summer of 2021, a data collection effort was conducted at GTRI, wherein multiple tag designs were placed at the bottom of Casey Pond at GTRI, as well as in the diving pool at Kraken springs. High resolution SONAR tracks have been acquired, containing the backscattered responses from these AID tags.

As part of the GRIP internship, student(s) will investigate signal processing, and machine learning methods to detect and classify these AID tags from the data available, quantifying detection accuracy by means of relevant performance metrics (e.g. mAP scores). Should time permit, students will also aim to theoretically characterize detectability ranges, and AID tag performance limitations using the available datasets, with the option to collect or simulate additional SONAR data for benchmarking using SONAR hardware.

It is desirable that interested students have some exposure to machine learning, probability and statistics, image classification, digital signal processing and acoustics. The ability to efficiently program in Python and Matlab is necessary for this role.

Citizenship Required:

US-Citizens Only

In-Office Time Required: 100%

Competencies:

Electrical Engineering, Mechanical Engineering, Computer Science

Project Description:

Thruster coatings are crucial for spacecraft as they provide thermal protection, prevent overheating during propulsion, and offer resistance against corrosion in the harsh space environment, ensuring the structural integrity and longevity of the spacecraft's propulsion system. This project aims to gain insights into the characteristics of thruster coatings tailored for future USSF missions, evaluating their durability under electron irradiation of the simulated Geostationary Orbit (GEO) environment. This project will be executed in close collaboration with the Kirtland Air Force Research Laboratory (AFRL), underscoring the significance of a cooperative approach in harnessing expertise and resources.

GRIP students will engage in a comprehensive exploration of the morphological characteristics of thruster coatings deposited at the AFRL. Employing cutting-edge methodologies, the coated coupons' surfaces will undergo meticulous evaluation through a suite of advanced techniques, including atomic force microscopy (AFM), scanning electron microscopy (SEM), and optical microscopy. Furthermore, the investigation will delve into the elemental composition of the coatings, utilizing a state-of-the-art laser-based elemental analyzer in conjunction with a 3D Keyence digital optical microscope. Learning objectives of the proposed project include:

1. Training of interns on the usage of the respective equipment (AFM, 3D microscope, etc);
2. Material characterization measurement of the respective samples;
3. Compiling a comprehensive report that delineates the experimental results and actively participating in effective communication with AFRL researchers. We expect to include the findings of research performed by Summer interns in a follow-up conference or journal publication prepared by the GTRI scientists.

Citizenship Required:

US-Persons Only

In-Office Time Required: 100%

Competencies:

material science, physics, optics, chemistry

Project Description:

The aim of Pufferfish is to create a model and simulation environment that synthetically produces LiDAR imaging of the ocean surface and water column. Characteristics of the GTRI LiDAR system are combined with knowledge of the ocean environment, ocean physics and phenomenology, and optical beam physics/radiative transfer in the water column to calculate collections of point clouds. The project covers topics which include fluid mechanics/physics, oceanography, and radiative transfer modeling.

Last year's GRIP students successfully established a working model and simulation environment of LiDAR point clouds that could capture rudimentary internal gravity waves. Retrieval algorithms were also developed to extract fundamental information about the ocean state and internal gravity wave structure under various environmental conditions. The prospective students would continue and expand upon this work, in particular, by:

1. applying these retrieval algorithms to real-world LiDAR data sets
2. increase the fidelity of our gravity-wave modeling capabilities by adding non-linear effects
3. combine/couple global ocean and/or internal gravity wave models into our modeling and simulation framework

The students should have experience with programming, preferably in MATLAB, Python, and/or C/C++. Experience with finite element modeling or computational fluid dynamics is also desired.

Citizenship Required:

US-Citizens Only

In-Office Time Required: 0%

Remote Option: Yes

Competencies:

Physics, fluid mechanics, math, computer Science, optics, electrical engineering

Project Description:

Robotic Manipulation in semi- and un-structured environments, e.g., when the exact appearance and pose of the objects is not known ahead of time, is a fundamental problem standing on the path to a wider deployment of robots. The approaches that attempt to address this problem rely on Machine Learning algorithms in order to interpret the scene and describe the objects. Due to the amount of training data required and difficulty of generating dense 3D annotations for robotic tasks, simulators are often used to create training datasets for Machine Learning. However, the difference between simulated and real data (sim2real gap) limits the performance of image-based Machine Learning models trained on synthetic data when applied in real-world.

This project will leverage recent Generative Image models, such as Stable Diffusion, in order to develop automatic modification strategies for synthetic images based on user-engineered text queries. Example modifications include generation of alternative object textures and backgrounds for domain randomization to improve generalization of trained models to real data. Development will be benchmarked in relation to a baseline synthetic dataset and tested on real dataset and on the physical robot arm.

Citizenship Required:

No

In-Office Time Required: 0%

Competencies:

Computer science, machine learning, robotics

Project Description:

Project Objective:
This project aims to advance technology and cultivate a new generation of skilled researchers and innovators in the field of optical engineering and imaging sciences. Students will delve into the multifaceted remote sensing applications of a Scheimpflug Camera. Each student will be tasked with understanding and expanding upon existing GTRI research to explore a specific application: passive rangefinding, topographic mapping, vision for autonomous vehicles, photogrammetry, or optical turbulence measurements. Under the guidance of their mentor and GTRI colleagues, they will evaulate their systems' capabilities in relation to real-world needs. Through hands-on experience in modeling, simulation, and prototyping, the students will not only contribute to the advancement of this technology but also gain valuable skills and knowledge in their field. This project is designed to challenge and inspire the next generation of students interested in optical engineering by encouraging creative problem-solving and technical proficiency in a collaborative learning environment.

Skills Required from Students

1. Technical Knowledge: Understanding of optical systems, camera technology, and image processing.
2. Research Skills: Ability to conduct literature reviews and synthesize information from scientific papers.
3. Problem-Solving: Innovative thinking to overcome technical challenges and adapt to new concepts.
4. Technical Proficiency: Familiarity with python for modeling and simulation
5. Hands-On Experience: Skills in prototyping and experimentation

1. Interdisciplinary Exposure: Opportunity to work at the intersection of physics, engineering, computer science, and applied mathematics.
2. Research and Development: Learn how to transform theoretical knowledge into practical applications and prototypes.

Citizenship Required:

US-Persons Only

In-Office Time Required: 100%

Competencies:

optical engineering

Project Description:

In the rapidly evolving landscape of electronic warfare and cyber operations, this project aims to investigate the viability and effectiveness of employing a low-cost jamming solution (Low Cost Kitten) to enhance Cyber Electromagnetic Activities (CEMA) within contested Radio Frequency (RF) environments. The Low-Cost Kitten (LCK) is a cost-effective variant of the Angry Kitten (AK) that enables dynamic jamming of RF transmissions using simple and cheap COTS technology. With only small modification, the LCK can operate within the radio frequency band of ground-based army systems and provide electronic warfare (EW) capability to army units that currently don't have this capability. With the increasing reliance on wireless communication and electronic systems, the electromagnetic spectrum has become a critical domain for military operations. This research will encompass the development, testing, and evaluation of the LCK in CEMA operational environments, considering factors such as affordability, portability, and adaptability for various operational scenarios.

Key objectives include:

• Evaluating the cost-effectiveness and technical capabilities of the proposed jamming system.
• Assessing its ability to operate within contested RF environments while avoiding interference with friendly forces.
• Investigating the impact of jamming on friendly/adversary electronic warfare capabilities and communication networks.
• Analyzing the jamming system's adaptability and ease of deployment in diverse operational settings.

The outcome of this project will provide valuable insights into the feasibility of incorporating low-cost jamming technology into CEMA operations, potentially expanding the U.S. Army's capabilities in electronic warfare. Moreover, it will contribute to enhancing the Army's competitive edge in modern conflicts where control of the electromagnetic spectrum is of paramount importance. Further, this project seeks to expand the current EW Test and Evaluation (T&E) portfolio of GTRI’s ELSYS/TEN Division, by investigating the feasibility of using USAF aircraft developed technology for US Army infantry capabilities.

This feasibility study will be conducted through a combination of simulations, laboratory testing, and field exercises, with the aim of delivering actionable recommendations for further development and integration of the proposed low-cost jamming solution into the Army's CEMA arsenal as a training/development asset.

Citizenship Required:

US-Citizens Only

In-Office Time Required: 100%

Competencies:

Computer Engineering, Computer Science, Cybersecurity Engineering, Electrical Engineering

Project Description:

Students will measure and analyze various properties of model-scale, internally-mixed dual-stream jets. This work will include measuring acoustics in a large anechoic chamber, visualizing jets using high-speed schlieren photography, obtaining thrust measurements, and potentially utilizing Particle Image Velocimetry (PIV). These measurements are helping to shape how the nozzles of future supersonic/hypersonic jet engines are designed and will play a key role in reducing the noise made by supersonic aircraft. These data are highly sought after by CFD and modeling teams across the country, including groups at Stanford, Penn State, and University of Illinois. Students are expected to be comfortable working with their hands, as changing the experimental set-up, calibrating and aligning microphones, and fine-tuning schlieren captures will be common occurrences.

The students working on this project will have the option of also working closely with Professor Ahuja of the AE School, who is also the chief of the Aerospace and Acoustics Technologies Division (AATD). This this option, in addition to the internship, students wishing to also get a grade for research can register for a special project during the summer or as an option during the following Fall. Professor Ahuja is willing to take in up to four students who can complete research at GTRI during the summer and then either finish a report and receive a grade during the summer or continue meeting him regularly the following semester to complete a research report for a grade. We envision allowing student teams who select this option to spend up to 2 hours each day on innovative and disruptive projects for this purpose. These students will need to submit an Undergraduate Research Permit for AE4699 or equivalent and have it approved by Dr. Ahuja prior to the semesters starting.

Citizenship Required:

US-Citizens Only

In-Office Time Required: 100%

Competencies:

Signal processing, MATLAB, Experimental Acoustics

Project Description:

A digital twin (DT), or virtual representation of a physical entity or system, is typically constructed with digitized data collected from representative physical entities using sensors that capture the observable behavior of the entities and then integrate these data into a digital emulation.

Defense and industry have only explored and adopted the concept of DTs to a limited extent, mainly involving the emulation of physical assets and production processes. Other spaces, however, have extended the roles of the DT concept; information-processing and cybersecurity use DTs to emulate human adversaries who conduct cyber attacks, while health and fitness spaces use DTs as personal assistants that track observable markers of cognition, health, and performance.

Although there is often debate in diplomatic, executive, and academic venues regarding the ethical implications of emerging artificial intelligence capabilities, there is a relative paucity of empirical investigation exploring the ethical and functional boundaries of digitizing a human mind, although what research has been accomplished to date suggests that AI has immense potential to help improve the quality of human experience, particularly in challenging, high-risk scenarios in which this pairing of AI and human could benefit the human condition or potentially even save a human life.

In this project, we will explore the feasibility and ethical implications of the human digital twin (HDT) concept to emulate human operator cognition in high-demand, high-risk command, control, and communications (C3) environments to optimize the human decision process, enhance crew performance, and lower operational risk. Using a human-centered artificial intelligence (HAI) approach, we will design a preliminary HDT concept to be implemented independently or teamed with a human operator in a range of possible scenarios to include mission planning, training, and execution.

Citizenship Required:

US-Citizens Only

In-Office Time Required: 100%

Competencies:

artificial intelligence, machine learning, neural networks, data science, computer science, cognitive science, psychology, digital twin

Project Description:

The combination of virtual reality and the remote control of semi-autonomous ground based quadruped robots offers many benefits in searchand rescue operations. The project will use COTS equipment to develop a deployable pack of VR controlled quadruped robots. The project will develop the communications and control links need to allow the user to control and direct the robots using a VR headset. The project will also investigate the sensor array required on the robots so that the user has a seamless 360 degree situational awareness of the remote surroundings. The sensor array will include vision and audio sensors that when combined will create a dynamic realistic surround sound VR experience.

Citizenship Required:

US-Citizens Only

In-Office Time Required: 50%

Competencies:

Computer Engineering, Computer Science, Electrical Engineering

Project Description:

Thousands of spacecraft measurements are collected by electro-optical and radar systems every night from locations around the globe from sensors with distinct hardware qualities and varying weather conditions. These datasets must be organized to provide satellite operators with interpretable information that can be acted upon decisively in a major event. This was proven on Nov. 15, 2021 when Russia tested an anti-satellite missile, spawning a debris cloud that forced the International Space Station to maneuver to avoid collision. In this project, we will create an autonomous forecasting system using State Space Modeling (SSM) approaches that are augmented by Machine Learning (ML).

A SSM is rooted in filtering theory and is described by:

1. a state vector tracked over time
2. a state-transition equation defining the state vector velocity
3. an observation model specifying the conditional probability of observations given the state vector

This model is potentially advantageous because

1. it allows incorporation of known dynamics
2. it can be trained/updated in real time
3. it is highly interpretable due to the tracking of covariance

Unfortunately, traditional SSMs are fitted on each time series individually, and therefore cannot infer shared patterns from a dataset of similar time series (i.e. satellites in the same orbital regime). We will explore an approach for using ML models to learn a global mapping for a SSM across a batch of similar time series: Deep SSMs (Rangapuram, 2019). We believe that this approach can significantly enhance the fusion of diverse data streams of satellites, ultimately providing an interpretable alert system for satellite operators. This approach will be proven both with our in-house 3D simulation software as well as a dataset provided by the Joint Task Force Space-Defense (JTF-SD). Students will gain substantial expertise in working with ML models operationally and with simulating datasets to train ML models.

Citizenship Required:

US-Citizens Only

In-Office Time Required: 100%

Competencies:

Mathematics, Physics, Computer Science, Data Science, Applied Mathematics

Project Description:

Overview:
Event cameras revolutionize optical systems by asynchronously processing brightness changes, boasting >120dB dynamic range, 1-100Âμs temporal resolution, and >2 Giga-events-per-second bandwidth. This technology excels in space awareness, chemical imaging, microscopy, autonomous driving, and has potential in remote sensing and neural imaging.

Challenge:
Their adoption is hindered by the absence of practical databases, limited simulators, and unclear data statistics like noise, affecting the development of generalizable algorithms.

Innovation:
This GRIP project introduces a dual approach. Firstly, it funds a Python package development for generating, characterizing, and processing event-based data, accelerating technology advancement. It aims to incorporate self-supervised deep learning for future algorithmic development. Secondly, it supports creating an in-house event-based camera platform to study the relationship between light conditions, camera parameters, and resulting event streams. A practical dataset of natural events will also be developed for algorithm testing and characterization. This project positions EOSL at the forefront of event-based technologies, simplifying the innovation process in optical platforms for various applications.

Experience Gained:
Software Development: Hands-on experience in Python package creation.
Data Management and Analysis: Skills in building and maintaining a specialized database.
Event-Based Imaging Insights: Deep understanding of this innovative technology.
Practical Machine Learning Application: Experience with real-world AI and algorithm integration.
Problem-Solving in Context: Tackling industry-relevant technical problems.
Project Management: Skills in overseeing and coordinating a tech project.
Presentation Skills: Ability to convey technical information effectively

Citizenship Required:

US-Persons Only

In-Office Time Required: 100%

Competencies:

Python Programming, Data Structures Knowledge, Machine Learning Basics, Image Processing, Optical/Physics Simulations

Project Description:

The "Multi-agent AI Applications for DoD" project aims to develop methodologies to automate useful assessments based on information maintained within a knowledge management system, producing insights that can be embedded into the views of the knowledge management site. The project will leverage large language models and predictive modeling to provide added value to decision-makers.

Large language models can be used to extract information from the knowledge management system, generate insights from the data, and identify patterns and trends that would be difficult or impossible for humans to spot. Predictive modeling can be used to forecast future events and outcomes and to develop recommendations for decision-makers.

By combining large language models and predictive modeling with multi-agent AI, the project can develop a system that can help DoD decision-makers to be more informed and effective in their work.

Citizenship Required:

US-Citizens Only

In-Office Time Required: 100%

Competencies:

Computer Science, Data Science, Mathematics, Software Development

Project Description:

We will develop a novel compact, sensitive, low-cost and flexible infrared sensor for applications including chemical/bio detection, optical communications, IR imaging and UAV localization etc. This new device is based on cutting-edge photosensitive materials and fast electron-transporting thin films, which enables low-cost solution processing, physical flexibility or conformal coating on bendable subjects.

Advanced standoff systems that detect and classify chemical and biological contaminants are critical components of any chemical/biological defense system needed to maintain homeland security. While hyperspectral imaging provides the required performance, its widespread deployment is currently limited by the cost, yield, and reliability of the infrared (IR) sensor technology used. In addition, compact and sensitive IR sensors are required as a complement or replace RF functionality in applications including optical communication, UAV localization and imaging, surveillance and machine vision.

Most commercial short to mid-wave IR sensors or photodetectors are based on crystalline inorganic semiconductors, such as silicon, InGaAs, HgCdTe and Ge. However, these materials are complex and expensive in manufacture, and require high driving voltage and cooling to ensure high efficiency. Also, they are physically rigid and difficult to be used in applications requires light weight, device flexibility and distributed sensing/imaging capabilities.

Using solution-processable materials for IR device offer great advantages over CMOS processing: such as low fabrication cost, less physical rigid, easy to form contour shape, or onto flexible textile substrate. Moreover, the quantum dot materials show unique properties including tailorable band gap which means tunable emission, efficient light absorption, and high charge-carrier mobility.

Citizenship Required:

US-Persons Only

In-Office Time Required: 100%

Competencies:

Materials science and engineering, Chemical & biomolecular engineering, Electrical engineering, Aerospace engineering, Applied physics, Mechanical engineering

Project Description:

Vapor cells filled with Rubidium or Cesium are used in many important applications, such as portable atomic clocks, magnetometers, and Rydberg rf-sensors. However, their production is still a bespoke process.

Vapor cell production involves a few key challenges:

• fabrication of a vapor cell chamber
• integration of optical windows
• loading of chambers with process gas
• creating full hermetic sealing of the chambers to contain the process gas

4" silicon wafers are etched with MEMS fabrication processes to create individual die consisting of 2 vapor cavities and gas lines connecting the cavities.

We are exploring micro-pills as a vehicle for dispensing rubidium (Rb) vapor but are also investigating additional methods of dispensing vapor such as liquid-phase solutions. For micro-pills, one of these cavities acts as a holding cell for an un-activated Rb micro-pill and the other operates as the functional cavity to be probed by lasers once the micro-pill is activated. Micro-pills are loaded into the storage cavity before the wafer is hermetically sealed with additional MEMS processing. The micro-pills are activated with heat, releasing Rb vapor into the functional cavity, resulting in a ready-to-use vapor cell.

There are no existing automated methods designed around the dispensing of micro-pills onto wafers. Micro-pills are 1.08mm in diameter, 0.6mm in height, making them difficult to manipulate by manual means. Additionally, micro-pills need to be specifically oriented when placed inside the loading cavities prior to hermetic sealing. Liquid-phase dispense methods involve the use of inert-gas environments or requiring processing under vacuum. Different tool designs will need to be created to consider the medium of vapor delivery.

This project seeks to create a cost-effective automated method of loading vapor-carrying micro-pills and other vapor-carrying mediums into MEMS fabricated wafers which can be integrated into the fabrication workflow of a vapor cell at scale.

Citizenship Required:

U.S. Citizens Only

In-Office Time Required: 100%

Competencies:

Mechanical engineering

Project Description:

The StellarNAV framework is a spacecraft navigation method that uses the perturbations in observed starlight to autonomously navigate spacecraft. One implementation of the StellarNAV framework uses relativistic perturbation of starlight in apparent direction (stellar aberration) from multiple stars to calculate inter-star angles. These measurements are used to calculate the spacecraft velocity, which in turn is used as the basis of the spacecraft navigation filter.

While the technology to accurately measure stellar aberration exists, it has not been implemented in a system that both uses multiple optical instruments and needs to meet the size and compatibility constraints of a spacecraft navigation system. The three main challenges to implementing this framework are related to the performance and size of a single optical instrument, the alignment between optical instruments, and pointing and vibration requirements.

This project aims to progress the development of this framework by performing trade studies, drafting requirements, and performing the initial design of the optical, opto-mechanical, electrical, and/or software aspects of the StellarNAV platform. Each student will be able to contribute to the project based on their own interests and areas of expertise. Students will identify, understand, and address challenges related to the implementation of the system and, by the end of the project, deliver a preliminary design of their chosen aspect of the platform.

Citizenship Required:

US-Persons Only

In-Office Time Required: 100%

Competencies:

Optical Engineering, Mechanical Engineering, Electrical Engineering, Computer Engineering, Aerospace Engineering

Project Description:

Students will design and create technology to enhance communication and machine learning (ML) tasks across systems such as IoT, smart grids, or satellite communication, focusing on overcoming the diversity and complexity in digital ecosystems. Current communication methods rely on a perfect understanding of the full network. This is not a reasonable assumption in an open world environment. For instance, in the context of satellite communication, it's unrealistic to always expect complete network information for transmission to all satellites. When we lack the full understanding of the network, capabilities can be assured using sheaf theory, a mathematical framework from geometry. ML techniques can then be deployed across the system for downstream tasks such as detecting cyber attacks.

The project aims to tackle the challenges posed by the heterogeneity of modern digital systems, developing advanced communication protocols and ML algorithms specifically designed for seamless integration of various elements of smart infrastructure. The application of sheaf theory will enable a deeper understanding of the structural properties of these networks, leading to more efficient and robust systems, and allowing for ML models to be trained and deployed on the network.

This technology has the potential to significantly impact the current cyber environment through the development of more efficient, robust, and intelligent digital infrastructure. This can lead to significant improvements in areas like energy management, resource allocation, and system resilience against faults or attacks.

Citizenship Required:

No

In-Office Time Required: 0%

Remote Availability: Yes

Competencies:

Applied Mathematics, Computer Engineering, Computer Science, Discrete Mathematics

Project Description:

The goal of this project is to research and develop a compact recovery system for CubeSat platforms that does not rely upon pyrotechnics for activation. The recovery system will focus on parachute deployment at hypersonic speeds, as the CubeSat falls to the earth. The goal of pyrotechnic-free deployment is to avoid the necessity of a 2-3 fault tolerant containment system for black-powder actuators, enabling launch of lightweight CubeSats from the International Space Station.

The system will focus on a dual-event recovery, which involves deploying a drogue parachute prior to the main parachute. Pursuing a dual-event recovery adds complexity and cost, but ultimately decreases recovery time and drift. NASA has previously explored a pyro-free recovery system in 2019 with MIRCA, a recoverable CubeSat platform. The MIRCA was tested and never recovered due to a parachute deployment failure. These lessons learned can help drive the student design considerations.

Students will fully characterize the design, and iterate through prototypes. At the end of the program, the students will perform either a drop test or a balloon launch to test the recovery system.

Citizenship Required:

US-Citizens Only

In-Office Time Required: 100%

Competencies:

Aerospace engineering, mechanical engineering

Project Description:

The goal of this project is to measure the characteristics of supersonic flow through a miniature tunnel and support the development of a scramjet experimental facility. Reliable propulsion is one of the challenges of hypersonic flight. Therefore, this tunnel is being developed to support research in the development of scramjet engines.

Flow diagnostic experiments are imperative to understand the conditions being experienced at the inlet to a combustor and throughout the features of the combustor. This work will employ PIV and high-speed Schlieren, to investigate flow characterization such as flow uniformity and shock locations.

Citizenship Required:

US-Citizens Only

In-Office Time Required: 100%

Competencies:

Aerospace engineering, mechanical engineering

Project Description:

The student interns for this GRIP project will research, develop, fabricate, test and deploy a multi-functional agricultural robotics platform for urban farming in metro Atlanta. Tasks such as harvesting, phenotyping, pest management and yield prediction will be explored and implemented. The platform will serve educational, research and food production purposes.

In addition to applying their engineering skills, the project will require team members to engage with the community and other stakeholders to site, integrate and commission the agricultural robotics platform.

Citizenship Required:

No

In-Office Time Required: 100%

Competencies:

mechanical engineering, computer science, bioengineering, electrical engineering, other

Project Description:

Clean hydrogen, produced with no carbon emissions, has recently received considerable attention as a key element in the transition to clean energy, particularly for hard-to-decarbonize sectors such as heavy-duty transportation and industrial applications. It also plays a crucial role in enabling long-term energy storage for a clean electric grid. Additionally, numerous military organizations have expressed strong interest in hydrogen fuel cell technology, exploring strategies for its widespread application in military devices, including advanced weapon systems, future combat systems, unmanned aerial/ground vehicles, soldier portable power, and silent camp/watch capability.

Water electrolysis is considered the most promising pathway to produce clean hydrogen with zero emissions. While water electrolysis systems are commercially available today, they still face challenges in terms of affordability, durability, and efficiency. Much research and development is currently underway in both public and industry sectors to reduce the production cost of clean hydrogen to $1 per kg within a decade (currently $5 per kg). Designing an efficient electrolysis cell and optimizing system operation parameters (e.g., temperature, current, flow rate, etc.) are essential for the reduction of clean hydrogen production.

This research program is designed for undergraduate and graduate students to gain insights into clean hydrogen production technology through water electrolysis. Through the assigned research tasks, students will not only grasp the basic concepts of clean hydrogen production via water electrolysis but will also acquire hands-on experience in fabricating lab-scale water electrolyzers and operating systems to produce clean hydrogen. The outcomes of this project can also contribute to forming a crucial feedback loop, identifying barriers, and validating progress toward overcoming these barriers in state-of-the-art water electrolyzers.

Citizenship Required:

No

In-Office Time Required: 100%

Competencies:

Chemistry, Chemical Engineering, Material Science, Mechanical Engineering