In the D2K Capstone program, interdisciplinary teams of students (advanced undergrads, professional master's students, and Ph.D.) work on a semester-long real-world project sponsored by our D2K Affiliate Members.
Fall 2024 Projects
CHRONICLE
Team Name: Chronicle Outages
Power Outage Analysis | Houston, we have a problem. Weather events 
cause persistent, long power outages in Houston. In turn, the Houston Chronicle and our team aim to answer the question, “What factors affect the frequency and duration of power outages in CenterPoint Energy’s electrical circuits"?
Student Team Members
- Sejal Gupta
- Shreya Challa
- Yuxin Yan
- Jacob Lapp
- Caroline Hashimoto
- Daniel Zhao
Sponsor and Mentors
- Sponsor: Matt Zdun
- D2K Fellow, PhD Mentor: Tyler Bagwell
- Faculty Mentor: Arko Barman
FE&P
Team Name: The Kilowatts
Powering a Greener Future: Detecting Energy Overuse at Rice University | This project aims to 
optimize building energy consumption at Rice by preventing wasteful energy overuse events. Through the Office of Sustainability and Facilities and Capital Planning department at Rice University, our project leverages available utility meters and weather data to pinpoint energy waste at the building level and develop a low-cost solution to predict consumption.
Student Team Members
- Robert Hunt
- Zac Ambrose
- Masha Zaitsev
- Sofia Lakhani
- Junie Wei
- Zhen Qin
Sponsors and Mentors
- Sponsor: Terie McClintock, Keaton Kinstley, Johnny Pickle
- D2K Fellow, Ph.D. Mentor: Roy Phillips
- Faculty Mentor: Dr. Xinjie Lan
HFD DEMOGRAPHICS & ENVIRONMENT
Team Name: The Dalmations
Environmental and Demographic Factors to Houston Fire Department Emergency Response Demands | Houston Fire Department
responded to over 400,000 emergency incidents in the past fiscal year, which included emergency medical services, motor vehicle collisions, hazardous materials releases, false alarms, and fire rescues. Identifying environmental and demographic favors that contribute to service demands may enable Houston Fire Department to develop mitigation strategies and improve resource allocation.
Student Team Members
- Tina Li
- Kaylah Patel
- Namita Jatkar
- Vivian Le
- Tom Luan
Sponsors and Mentors
- Sponsor: Leonard Chan
- D2K Fellow, Ph.D. Mentor: Nhi Le
- Faculty Mentor: Dr. Arko Barman
HFD RAINFALL
Team Name: Team Flood Fighters
The Effects of Rainfall on Emergency Response Times for the Houston Fire Department | As the Bayou City, Houston
experiences severe rain and flooding incidents on a regular basis, which can hamper the ability of the Houston Fire Department to quickly and effectively to over 400,000 annual calls. Utilizing data from both HFD and the Harris County Flood district, this project identifies the causes and effects of rain delays to help HFD more effectively respond in inclement weather.
Student Team Members
- Nico Motta
- Lauren Yu
- Jared Boyd
- Cunyi Mao
Sponsor and Mentors
- Sponsor Mentor: Leonard Chan
- D2K Fellow, PhD Mentor: Jiaming Liu
- Faculty Mentor: Dr. Xinjie Lan
HGO
Team Name: The Operagoers
Demographic and Time-Dependent Factors to Houston Grand Opera Donor Conversions | Roughly 80% of 
Houston Grand Opera’s annual revenue comes from donations, with only 20% being accounted for by ticket sales. Identifying demographic and time-related factors that contribute to an opera attendee’s likelihood of becoming a donor may enable Houston Grand Opera to develop marketing strategies and increase annual revenue.
Student Team Members
- Gal Kadmon
- Yurie Han
- Amy Lam
- Junyao Ren
Sponsor and Mentors
- Sponsor Mentor: Grace Tsai, Claire Padien-Havens
- D2K Fellow, PhD Mentor: Janet Fu
- Faculty Mentor: Dr. Xinjie Lan
KINDER HERC
Team Name: Team Kinder Herc
Bridging the Gap: Making Education Data and Research Accessible to Texas School Districts | Our project analyzes 
Texas school district data, merging demographics, STAAR test results and attendance rates into a unified dataset. We use PCA and autoencoder models to derive a district similarity metric. Visualizations make insights accessible, enabling districts to make informed decisions.
Student Team Members
- Victor Xie
- Melissa Mar
- Nate Lee
- Anu Jain
- Ananya Kapoor\
- Wensheng Chu
Sponsor and Mentors
- Sponsor Mentor: Erin Baumgartner
- D2K Fellow, PhD Mentor: Mauro Florez
- Faculty Mentor: Dr. Xinjie Lan
NABORS
Team Name: Team Nabors
Nabors: Analyzing and Optimizing Rig Inventory Processes | Nabors Industries spends over 1 billion dollars 
every year on excess inventory for its oil rigs. This project has two primary objectives: first, to analyze and model data to uncover unusual purchasing patterns, and second, to develop a purchasing recommendation model leveraging stochastic inventory processes and graph algorithms to optimize procurement decisions. Our ultimate goal is to streamline inventory management and significantly reduce unnecessary spending, driving both operational efficiency and cost savings.
Student Team Members
- Melody He
- Lucy Hu
- Audrey Pizzolato
- Tian Wu
- Faith Zhang
Sponsors and Mentors
- Sponsor Mentor: Gautham Kanaparthy
- D2K Fellow, PhD Mentor: Yihua Xu, Kevin McCoy
- Faculty Mentor: Dr. Lan, Dr. Shaw
NASA POSE
Team Name: SpaceCats
SpacePose: A Transformer-Based Cross Encoder Model for Spacecraft Pose Estimation | Inspector spacecraft 
are specialized vehicles designed to assess and repair damaged spacecraft in space. To ensure precise docking with the target vehicle, these inspectors require accurate pose estimation of the damaged spacecraft. We propose a novel spacecraft pose estimation model that can (i) reliably predict the orientation and position of various spacecraft types without prior knowledge and (ii) operate effectively in low-resource environments.
Student Team Members
- Christina Li
- Sheena Bai
- Eisha Hemchand
- Jae Jun Ku
- Anh Pham
Sponsors and Mentors
- Sponsor Mentor: James W. Berck
- D2K Fellow, PhD Mentor: Ananya Muguli
- Faculty Mentor: Arko Barman
NASA SEGMENTATION
Team Name: Galactic Maskers
Segmentation of an Unknown Spacecraft for In-Space Inspection | This project focuses on developing a 
real-time segmentation algorithm for in-space inspection of spacecraft using deep learning techniques. Given the diverse designs and configurations of spacecraft, the goal is to create a single model capable of accurately identifying any spacecraft, including previously unseen designs (i.e., unknown spacecraft). The project employs a pre-trained YOLOv8 nano model, which was fine-tuned using a combination of synthetic and real-world datasets. To align with deployment constraints, the model was optimized for CPU-only hardware, simulating the target operational environment. This algorithm aims to enhance NASA’s autonomous inspection capabilities, enabling improved spacecraft navigation and structural analysis under various visual distortions commonly present in space imagery.
Student Team Members
- Janhavi Sathe
- Jeffrey Joan Sam
- Naman Gupta
- Nikhil Chigali
- Radhey Ruparel
- Yichen Jiang
Sponsors and Mentors
- Sponsor Mentor: James W. Berck
- D2K Fellow, PhD Mentor: Janmajay Singh
- Faculty Mentor: Arko Barman
OCFR
Team Name: The Sci-Search Squids
ResQIR (Researcher Querying & Information Retrieval) | OCFR (Office of Corporate and Foundation Relations) 
at Rice University oversees the management of the university's relationships with corporate and foundation stakeholders. Our project aims to support OCFR by building a querying and visualization system that enables stakeholders to efficiently connect with faculty members based on shared research interests. This system will streamline access to expertise, fostering stronger collaborations and partnerships.
Student Team Members
- Aditi Raju
- Aldo Alvarez Gutierrez
- Jimi Abbott
- Keerthaa Golla
- Qizhi Li
Sponsors and Mentors
- Sponsor Mentor: Samantha Velez, Candice Pauley
- D2K Fellow, PhD Mentor: Lucy Liu
- Faculty Mentor: Arko Barman
TAMU
Team Name: SeeBirds
Deep object detection for waterbird monitoring using aerial imagery | The study of waterbirds 
(native to wetlands and freshwater ecosystems) and their populations often provides many signals and insights into broader ecological changes and environmental degradation. In colony islands where waterbirds are increasingly threatened by humans and rising sea levels, it becomes essential to be able to accurately assess waterbird nest locations and island dynamics in a cost and time effective manner. This project focuses on training object detection models to be able to locate and classify different species of waterbirds from aerial drone images taken along the Texas coast, provided by Texas A&M Corpus Christi as well as Audubon Texas. These machine learning tools will aid researchers in tracking foraging habits and nesting behaviors in order to better assess the health of coastal ecosystems.
Student Team Members
- Akshay Raj
- Jason Nguyen
- Will Miraglia
- Kaiyan (Kyle) Ma
- Sirui Hao
Sponsors and Mentors
- Sponsor Mentor: Anna Vallery, Richard Gibbons, Marissa Lamb
- D2K Fellow, PhD Mentor: Krish Kabra
- Faculty Mentor: Arko Barman