Lucas Young and Thomas Li, long-time friends, both worked as software engineers at Google. Young wanted to tackle a big challenge: improving night vision technology for the U.S. military.
Most night vision gear today is still analog. Goggles capture weak light using optical lenses and a chemical transformation, which can be costly—ranging from $13,000 to $30,000 from suppliers like L3Harris and Elbit America.
For years, the military has sought to digitize this technology, focusing mainly on the hardware side. A noteworthy example is the $22 billion budget for the Integrated Visual Augmentation System (IVAS), which was recently taken over by Anduril from Microsoft.
Young, who has a degree in computational photography, spent five years enhancing smartphone cameras. Li, on the other hand, has a passion for AI and computer vision.
One day, Young came across an intriguing paper titled Learning to See in the Dark, co-written by scientist Vladlen Koltun, now at Apple. The paper explored using AI for low-light imaging. However, at that time, the AI technology couldn’t yet deliver the required 90 frames per second for real-time use.
By 2024, Young noticed that advancements in AI chips had made 90 fps possible. He convinced Li to leave their jobs and start Deepnight, which quickly joined the Y Combinator winter cohort.
Their smartphone app captivates the military
The military was their primary target, but they needed a way in. Young found a conference where representatives from the U.S. Army’s night vision lab would be present.
He wrote a white paper explaining his vision: using software to solve night vision challenges. He distributed it at the event, including a copy to an army colonel who was interested enough to agree to read it. “It was just a casual encounter. I wore a T-shirt!” Young recalls.
The colonel found the concept intriguing and connected them with the US Army C5ISR Center for further discussions.
Eager to prove their idea, the duo built a basic night vision app for smartphones. They used a VR headset to hold the phone.
This simple prototype impressed the military and secured their first contract.
“In February 2024, just a month into Y Combinator, we received a $100,000 contract based on our app’s demo and the white paper,” Young shared.
Next, they had to formally showcase their technology. They flew to Washington D.C. to demo their software and advanced goggles to an audience of ten.
This presentation led to additional contracts. A year after launching, Deepnight had secured approximately $4.6 million in contracts from the U.S. Army, Air Force, and various companies.
Their success also drew investors. By the end of Y Combinator, they raised $5.5 million, including investment from Initialized Capital and notable angels like Kulveer Taggar and Matthew Bellamy, lead singer of the band Muse. Additionally, Koltun became an angel investor.
Deepnight’s focus is on software while collaborating with hardware producers, including goggle and military helmet manufacturers.
“Now, we can enable any device to see in the dark, be it cars, security systems, drones, or maritime equipment, all using a standard $50 smartphone camera. This makes our technology more accessible and cost-effective,” Young explains.
Check out this related article: Breaking the 15-Year Mold: What Pokémon Creators Have in Store for Fans
Source linkY Combinator,defense technology
