Skip to Content
Alumni connection

Ian Cinnamon ’14, Simanta Gautam ’17, Bruno Faviero ’15

Training AI to keep public spaces safer
Ian Cinnamon ’14, Simanta Gautam ’17, 
Bruno Faviero ’15
Ian Cinnamon ’14, Simanta Gautam ’17, Bruno Faviero ’15COurtesy of Synapse technology

A security checkpoint in Tokyo’s bustling Narita International Airport may not seem like ideal office space for a startup, but for Bruno Faviero ’15, Ian Cinnamon ’14, and Simanta Gautam ’17, cofounders of the computer vision company Synapse Technology, that’s exactly what it was for three months in 2018.

“We had two desks in between the x-ray machines, and we would work there for 12 hours, talking to the screeners, watching the system work, making tweaks,” Faviero recalls of this period of fine-tuning the artificial intelligence for their software platform. “I literally don’t think we could have been closer to our user.”

Faviero is chief operating officer of Synapse, which builds on research Cinnamon conducted while an undergraduate in MIT’s Department of Brain and Cognitive Sciences, showing that people working as TSA screeners at chaotic security checkpoints can be prone to fatigue and distraction. The three alumni founded the company after realizing that they might be able to improve the process with AI.  

Faviero explains that Synapse’s deep-learning algorithm is trained on millions of images to recognize dangerous objects, no matter how they’re positioned inside luggage. The system consists of a server and monitor—powered by AI software—that are added to an existing x-ray machine. The monitor displays what the AI is “seeing” as bags are scanned, and alerts the machine’s human operator to potential threats by boxing suspicious areas of the images.

Synapse’s algorithm has become sufficiently adept at identifying guns and knives that humans—who examine every image, including those with AI annotations—can focus on spotting a smaller list of objects, reducing cognitive load and fatigue. In a simulation experiment, untrained human screeners assisted by the technology detected significantly more threats than trained screeners working without it. The founders hope that as Synapse’s AI improves and learns to detect more items over time, detection of most prohibited items can eventually be automated, allowing airports to keep up with increasing passenger volume.

Synapse started in 2016 in Palo Alto, California, where the team, hungry for data to train the AI, created a replica security checkpoint in the garage of a rented house. There they performed thousands of screenings on bags of all shapes and sizes, periodically inviting gun store owners to bring firearms for tests. At the same time, the cofounders began building partnerships with airports, eventually piloting their technology in Tokyo and Stockholm. Since then, the system has become commercially available; it is now being used at airports in San Jose, California, and Osaka, Japan, as well as in France and the Netherlands.

Cinnamon, who serves as Synapse’s president, holds an MBA from Stanford University and has experience founding and leading startups: he was a founder of the incubator Superlabs (acquired by Zynga) and led the Immunity Project, a nonprofit developing a vaccine for HIV, through Y Combinator. For Gautam (the chief technology officer) and Faviero, both computer science majors, Synapse is their first startup, though neither is new to AI. Gautam’s research in industry and academia included a stint at NASA, where he integrated deep learning into remote sensing, while Faviero was an early employee of the analytics and machine learning company Kensho.

Faviero credits MIT with developing his entrepreneurial mind-set. “I don’t think I even knew what a startup was until I got to MIT,” he says. At the Institute, he became the director of the undergraduate entrepreneurship organization StartLabs. With several classmates, he also founded HackMIT, one of the largest student-run hackathons in the country.

Synapse is now developing an x-ray machine into which its AI system is fully integrated. The company is also expanding to serve courthouses, schools, prisons, and military bases. It has already tested its system at Florida’s Palm Beach County Courthouse and recently finalized a contract with the US Air Force, with more deployments to come.

“It’s been really interesting to learn about the broader security market,” says Faviero. “We’re bringing a lot of innovation to this specific domain.”

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.