Teen Develops Flood-Detecting CubeSat – IEEE Spectrum

Currently satellites, synthetic aperture radar, and GPS are used to collect data on flood damage, track the location of victims, and communicate with emergency responders. But technology failures and slow data transmission speeds lead to delays in response time, Merchant says. The increase in global flooding has intensified the need for more accurate and reliable methods.

Last year Merchant built what she says is a more effective way to track and collect data during floods: a small, inexpensive, standardized CubeSat integrated with artificial intelligence. The little satellites use a multiple of 10- by 10- by 10-centimeter units—which allows manufacturers to develop their batteries, solar panels, computers, and other parts as off-the-shelf components.

The CubeSat takes images of an area and uses pattern recognition to detect flooding, assess infrastructure damage, and track survivors.

Merchant presented her paper on the device at this year’s IEEE Region 3 annual conference, IEEE SoutheastCon.

“IEEE is a foundational part of my growth as a young researcher,” she says. “It turned engineering from my dream to reality.”

Building a CubeSat at MIT

Merchant says her interest in disaster response was sparked after learning that it can take several hours for emergency workers to receive satellite data.

Determined to find a faster method, she began researching technologies and discovered what CubeSats can do.

“CubeSats are very agile, scalable, and capable of forming constellations (multiple-satellite groups) that update data in nearly real time,” she says. “The idea that these small satellites—which fit into the palm of your hand—could deliver life-saving insights faster than traditional systems really inspired me to push the concept further.”

Last year Merchant and three of her classmates were accepted into MIT’s Beaver Works Build a CubeSat Challenge, where teams of up to five U.S. high school students were given eight months to develop a satellite capable of completing a space-based research mission.

Merchant’s team—the Satellite Sentinels—built a CubeSat powered by a convolutional neural network (CNN) that can identify heavily impacted flood zones and remotely collect data for disaster relief and environmental monitoring. CNNs analyze image data for pattern recognition.

Merchant was the group’s payload programmer and led the mission’s design and simulation efforts, which included planning, configuring hardware, and developing autonomous software and algorithms to manage the payload.

The team began by creating a 3D model of the device to visualize and refine the placement of its parts. The technology used—including a Raspberry Pi, multiple sensors, and a camera—was housed in a clear plastic cube.

The middle CubeSat was developed by Merchant and her team during the MIT Beaver Works Build a CubeSat Challenge. On the left is a commercial 1U CubeSat while on the right is a prototype of Merchant’s current design. Abigail Merchant

The device, which cost US $310 to build, weighs about 495 grams and was remotely connected to a laptop via Bluetooth during ground-based testing. The computer contains a machine learning algorithm—written by Merchant using Python—that analyzes collected images to detect flooding.

The CubeSat takes a high-definition image of its surroundings every 2 minutes and transmits it to the laptop. The satellite transfers up to 1,500 images daily and stores them on a 16-gigabyte SD card.

The algorithm then analyzes patterns, including changes in the water’s color and the image’s pixel density. When the algorithm detects flooding, the device can alert emergency responders.

“While many existing systems operate on multihour cycles, the CubeSat captures high-resolution images every 2 minutes,” Merchant says. “The system can then trigger alerts that are delivered to first responders via SMS or email.”

To test their system, Merchant and her team built a city model made of Lego blocks in an empty bathtub. They positioned the CubeSat over it, and it took images of the scene. They then added water and dirt to make it look more like a real flood. The CubeSat successfully transferred the images to the laptop, and the algorithm detected the flooding.

Out of 30 teams, the Satellite Sentinels placed third.

Continuing her work at Accenture

Merchant is continuing her research on flood-prevention technologies at Accenture in Richmond, Va., where she works remotely as a payload owner and designer for the company’s CubeSat launch team.

After the MIT program ended, Merchant decided to scale her project. She reached out to her former mentor Chris Hudson, the global technical lead in space cybersecurity at Accenture. He offered her an internship.

Merchant is working to make the transition from prototype to functional product but, she says, needs to overcome obstacles she encountered with her MIT project.

The main one was that the model struggled to detect flooding in variable conditions. It’s because the CNN model needs context, she says. Without it, the model can misinterpret complex visual cues. To fix the issue, Merchant trained the algorithm to spot flooding by identifying colors in individual pixels.

Transmitting images using Bluetooth worked in her bathroom, but it isn’t quite as useful when CubeSats are orbiting 700 kilometers above the ground.

“If you’ve used a Bluetooth headset before, you know it disconnects the moment you walk away from the device it’s connected to,” she says. “That isn’t going to work when the CubeSat constellation is in orbit.”

She suggested the Accenture team switch to SubMiniature Version A (SMA) antennas. The RF antennas connect to the CubeSats using an SMA connector.

“The development process has been one of the most formative experiences of my career so far,” Merchant says. “Working through the payload design and validation and meeting with these teams has given me so much experience, especially for my age.”

Her payload is expected to be launched early next year.

An aerospace internship at MIT

Merchant is an intern at the MIT Computer Science and Artificial Intelligence Laboratory, the school’s largest interdisciplinary lab, with 60 research groups. CSAIL is led by IEEE Fellow Daniela Rus, recipient of the 2025 IEEE Edison Medal.

The internship is remote, and Merchant conducts research in a laboratory at the University of Central Florida, in Orlando.

“IEEE is a foundational part of my growth as a young researcher. It turned engineering from a dream to reality.”

In that role, Merchant is focusing on cognitive cartography, a method for structuring complex information into semantic maps that reveal how ideas and concepts relate to one another. She uses embedding models, a type of machine learning that converts information into numerical representations. The embeddings allow computers to recognize similarities and relationships between concepts, even when they are described in different ways. The approach helps an AI product understand how ideas connect, rather than treating each piece of data as isolated.

“Being one of the youngest people in the lab is daunting,” Merchant says. “However, I’m really excited to learn from engineers and researchers who are working at the cutting edge of the field.”

She says she is hoping to attend MIT or Stanford.

The future of IEEE

Merchant was introduced to IEEE by Joe Jusai, former finance chair of the IEEE Orlando Section.

Her first personal experience with the organization happened in 2023 while she was conducting research for a science fair project. She was working on a robotic arm that could pick up items using an electroencephalogram and Bluetooth. The project was inspired by her grandmother, who suffers from mobility issues and was wheelchair-bound.

“I kept seeing IEEE mentioned in every regulation and standard I found,” Merchant says. When she learned about an upcoming Orlando Section meeting, she asked her mother to take her.

At the meeting, several members presented their research. Merchant asked Masood Ejaz and Varadraj Gurupur—the chapter chair and cochair—if she could discuss her science fair project.

“After presenting my work, IEEE quickly became a community that has shaped my understanding of what engineering can accomplish,” she says.

She felt on top of the world, she says, when she presented her paper about her CubeSat project at IEEE SouthEastCon.

“It’s one of those experiences that really changes you,” she says.

She is excited to become an IEEE student member when she starts college, she says. She also has her sights set on being elected as its president someday.

“I met Kathleen Kramer at one of my local IEEE events before she was elected IEEE president, and we spoke about my work,” she says. “After she was elected, I realized that I would love to become the president of IEEE someday.

“I hope one day that I can step into the same shoes as her and continue to help IEEE the same way it helped me.”