Harvard John A. Paulson School of Engineering and Applied Sciences

Harvard John A. Paulson School of Engineering and Applied Sciences Creating collaborative bridges across Harvard, educating the next generation of global leaders, and addressing society's greatest challenges.
(327)

Through teaching and collaborative research, The Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS) discovers, designs and creates novel technologies and approaches to societal challenges—in service to the world, the nation, and our community. We bridge disciplines, both within engineering and the applied sciences and beyond, to prepare broadly trained leaders, to advance foundational science, and to achieve translational impact.

As wildfires become more frequent, communities in Utah’s wildland-urban interface zones face growing risks to homes, inf...
08/28/2026

As wildfires become more frequent, communities in Utah’s wildland-urban interface zones face growing risks to homes, infrastructure and insurance coverage.

Nadia Colombi, S.M. ’22, Ph.D. ’25, is developing a wildfire risk model for the state that combines climate projections with building data to identify the features that influence whether a structure burns.

Her work builds on research in atmospheric chemistry and wildfire modeling at SEAS. Now a postdoctoral researcher at the University of Utah, Colombi is focused on turning technical research into practical information for policymakers, forest services and communities.

“I didn't ever want to give up the technical parts of the job that I do,” she said. “I like working in that space, and I like learning about new methods. But I also wanted to be able to interact with people more and feel like the work that I am doing can actually be useful.”

Her path from environmental science and policy to atmospheric chemistry reflects a broader goal: conducting rigorous research that can make a meaningful difference. https://bit.ly/3UqjVwW

Harvard SEAS researchers have demonstrated a promising new way to protect fragile quantum information using nothing but ...
08/27/2026

Harvard SEAS researchers have demonstrated a promising new way to protect fragile quantum information using nothing but mechanical vibrations — essentially extremely small sound waves.
https://bit.ly/4gz9uyG

When Raymond Liu’s brother lost his sight, Liu began thinking about how computer science could improve the lives of peop...
08/25/2026

When Raymond Liu’s brother lost his sight, Liu began thinking about how computer science could improve the lives of people with disabilities.

Now a Ph.D. student at SEAS, Liu has developed Mobilio, a smartphone-based navigation assistant for people who are blind or visually impaired. Inspired by earlier assistive technology research, Mobilio uses a device many people already have: a smartphone.

A recently published study in Nature Biomedical Engineering found that Mobilio improved navigation for blind and visually impaired participants. Liu plans to continue developing Mobilio while exploring other assistive technologies, including robotic exoskeletons.

“We’ve done the first outdoor validation study for Mobilio, and we're moving towards doing broader navigation studies where we go to participants' homes and evaluate Mobilio in areas around their communities,” he said. “But I am also thinking about other areas of assistive devices such as exoskeletons that could help people to walk. I’m interested in getting into areas that rely more on robotics.”

His long-term goal is to lead an academic research lab focused on using computer science and robotics to help people affected by injury or disease.

https://bit.ly/4xTNRzT
Learn more about Mobilio: https://bit.ly/4qwPn8V

Harvard engineers have developed a smartphone app called Mobilio that uses AI, built-in sensors and personalized audio b...
08/24/2026

Harvard engineers have developed a smartphone app called Mobilio that uses AI, built-in sensors and personalized audio beeps to give users with blindness or low vision continuous guidance and obstacle avoidance. Learn more: https://bit.ly/4qwPn8V

When the COVID-19 pandemic halted fieldwork in the Amazon, Matthew Stewart pivoted his SEAS Ph.D. from measuring emissio...
08/19/2026

When the COVID-19 pandemic halted fieldwork in the Amazon, Matthew Stewart pivoted his SEAS Ph.D. from measuring emissions in the rainforest to modeling and machine learning.

Mentored by Vijay Janapa Reddi, he dove into benchmarking datasets, algorithms, and hardware, and began thinking deeply about how AI systems intersect with policy and society. After completing his Ph.D., he stayed on as a postdoctoral researcher, focusing on the societal impacts of AI.

Today, Stewart is a lead AI researcher at Pelago Health, a digital substance use care provider. His work explores whether AI can help people access mental health and substance use support with strong safety guardrails and real clinical oversight. One of his current projects uses human-in-the-loop AI to help clinicians monitor for suicidality and intervene when warning signs appear.

“We're basically exploring this bridging factor of whether you can have an AI that allows people to get the clinical care they need, but with more safety guardrails and clinical oversight than a typical AI agent,” he said. "Because this space still doesn’t have a gold standard for how to do it, we're still trying to feel out exactly what makes the most sense. But we're definitely taking a very safety-cautious approach in the way we're building it.”

His interest in algorithmic accountability led to his award-winning paper “Beyond Explanation: Evidentiary Rights for Algorithmic Accountability,” which analyzed 168 court cases to show how access to evidence affects the ability to contest algorithmic decisions. His vision: a world where people can question and verify automated decisions, rather than simply being told to trust them. https://bit.ly/4cHsutp

08/13/2026

In the age of AI, how can we be sure that future autonomous systems — think ride share fleets, or groups of search-and-rescue vehicles — operate in a way we can trust?

Professor Stephanie Gil researches how autonomous robotic systems can be made safer and more trustworthy.

Learn more: https://bit.ly/4xtABSk

More than 80 countries face chronic drought, yet only a third have systems to act before a drought arrives rather than a...
08/11/2026

More than 80 countries face chronic drought, yet only a third have systems to act before a drought arrives rather than after. The cost of waiting is immense: the UN Food and Agriculture Organization estimates that every dollar invested in anticipatory action can save up to seven dollars in averted losses, while protecting lives and livelihoods before crisis conditions take hold.

Two students in Harvard’s Master in Design Engineering (MDE) program, Mitul Iyengar and Luke Fiorante (MDE ’26), set out to understand why more countries don’t act early—and what it would take to change that.

Their research took them to Mozambique, home to one of the world’s most advanced government-led drought anticipatory action systems. There, they mapped the full path from seasonal forecast to decision, speaking with more than 100 meteorologists, disaster managers, data scientists, and humanitarian practitioners.

Their core insight: the hardest part isn’t predicting drought. It’s deciding when a forecast is strong enough to justify action. Move too early and you risk mobilizing resources for a rainy season that turns out fine. Wait for certainty and the window to act closes.

The key is a trigger: the threshold at which forecasted risk becomes a decision to act. In Mozambique, that trigger is set collectively by a national technical working group spanning meteorology, disaster management, agriculture, water resources, and humanitarian agencies. The group analyzes decades of climate data and tests hundreds of thousands of thresholds against 25 years of drought history.

But this process largely lived in spreadsheets, scripts, and the working memory of a few experts—rigorous, but hard to see, teach, or replicate.

Iyengar and Fiorante’s answer is Ground State: a platform that turns trigger calibration from a specialist-dependent exercise into a repeatable workflow that national meteorological services can run independently.

Ground State builds on peer-reviewed methods developed by the World Food Programme and Mozambique’s National Institute of Meteorology (INAM), and surfaces them through an interactive map of a country’s own drought history. With a drag of a threshold slider, an analyst can instantly see which droughts would have been caught, which would have been missed, and where false alarms would have occurred—district by district, year by year.

The platform carries teams through the full annual cycle of calibration, in-season monitoring, and after-season review. It runs on a single machine, requires no cloud infrastructure, and is being prepared for release as open-source software so governments can inspect every line and fully own the system.

Iyengar and Fiorante co-designed Ground State with INAM meteorologists in Mozambique and analytics teams at the World Food Programme in Rome, and they are now in discussions to pilot the tool in new districts in the coming drought season. https://bit.ly/4g0IWG8

Research titled "A Random Matrix Perspective on the Consistency of Diffusion Models" received an Outstanding Paper Honor...
08/11/2026

Research titled "A Random Matrix Perspective on the Consistency of Diffusion Models" received an Outstanding Paper Honorable Mention at the 43rd International Conference on Machine Learning held last month in Seoul, South Korea.
https://bit.ly/4fZSD7K

08/05/2026

By day, James Arnold and David Pont work in the Harvard Biodesign Lab developing wearable robotics and AI tools to support upper-limb mobility and physical therapy. By night, they play in Luigi’s Fishtank, a band named after a piece of lab equipment at SEAS.

This summer, they brought both worlds together at “Abilities: Amplified,” a free community concert and tech showcase at the DCR Hatch Memorial Shell. Nearly 1,000 attendees came out to hear Luigi’s Fishtank, a jazz trio from the Berklee Institute for Accessible Arts Education, a cover band from PluggedIn, and Grammy-nominated blind artist Lachi.

Alongside the performances, attendees explored wearable technologies and assistive innovations from the Harvard Biodesign Lab, other Harvard labs, Boston-area biotech companies, and community organizations focused on disability and independent living.

“Abilities: Amplified” highlighted what happens when researchers, clinicians, musicians, start-ups, and disability advocates come together around a shared goal: expanding access, visibility, and community. The team hopes to bring the event back next summer. https://bit.ly/4wEeSqV

New study shows AI recommendations that can adjust to individual users could combat growing over-reliance on AI for deci...
08/04/2026

New study shows AI recommendations that can adjust to individual users could combat growing over-reliance on AI for decision-making support. https://bit.ly/4pUZe7T

Address

150 Western Avenue
Cambridge, MA
02134

Telephone

+16174952677

Website

Alerts

Be the first to know and let us send you an email when Harvard John A. Paulson School of Engineering and Applied Sciences posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Shortcuts

Share

Category