SPECTRAINTEL Lands Seed Funding to Advance Compact Hyperspectral Imaging Sensors for Defense and Space

SPECTRAINTEL, a startup developing ultra-compact hyperspectral imaging sensors, has raised a seed round from Kakao Ventures. 

SPECTRAINTEL is a deep tech startup building an integrated stack for material detection and identification — combining compact hyperspectral imaging hardware, industry-specific spectral databases, and analysis software. While conventional surveillance technology has focused on improving temporal and spatial resolution, SPECTRAINTEL adds wavelength data to the mix, aiming to determine what a target is, what condition it’s in, and whether it poses a threat, all through four-dimensional data.

Hyperspectral imaging captures dozens to hundreds of wavelength bands within a single image. Where standard RGB or thermal footage centers on shape and temperature, hyperspectral data breaks down the spectral pattern of each pixel to distinguish material type and condition. That makes it possible to detect hazardous material leaks, combustion flames, waste, structural degradation, or camouflaged objects from a distance and without physical contact — targets that are difficult to identify with the naked eye.

As its first product, the company is developing ASTRO-HSI, a hyperspectral adapter that attaches to standard astronomical telescopes. The device connects to existing telescopes to capture wavelength-specific images of celestial bodies along with pixel-level spectral data. SPECTRAINTEL is using the astronomy equipment market as an early entry point to quickly gather real-world user data, which it plans to use to refine its hyperspectral platform for industrial and defense applications.

With the new funding, SPECTRAINTEL plans to advance its compact hyperspectral imaging prototype and accelerate hiring of core talent. The company intends to expand its application scope step by step — starting with missile and projectile identification before moving into industrial non-destructive testing, environmental hazard monitoring, and foreign-material inspection in food and agriculture.

SPECTRAINTEL was founded by Dongho Yu, a physics student at KAIST. Since high school, Yu has pursued the question of how to identify distant targets through remote signals, conducting research spanning nuclear fusion plasma, exoplanets, projectiles, and space objects. His notable projects include a wide-field exoplanet detection system built on amateur astronomical data, a remote non-contact spectroscopy-based projectile identification system, and an optical surveillance system for detecting and tracking unidentified space objects. That body of work earned him recognition as a global finalist in NASA’s Space Apps Challenge — placing among the top 45 teams out of 11,511 entries — as well as an award from LIG D&A at a future aerospace academic conference.

Before founding the company, Yu spent about 16 months as first author on research at the Korea Institute for Basic Science’s Planetary Atmosphere Group, tracking changes in an unidentified absorber in Venus’s upper atmosphere. Rather than relying on machine learning, he developed his own interpretable mathematical model and presented the findings at the joint BACO-25 general assembly hosted by international meteorological, ice and ocean science societies. He is regarded as a rare founder who has personally worked through the entire pipeline — from physics and mathematical theory to AI, telescope and spectrometer hardware fabrication, and field measurement.

Youngmoo Kim, an investment manager at Kakao Ventures, said SPECTRAINTEL is a team seeking to transform the defense and space markets through ultra-long-range hyperspectral observation and analysis, having achieved both miniaturization and lightweighting through a self-designed device. He added that starting with an astronomical observation device that overcomes existing technical limits, the company is expected to expand into the defense market and deep space research — covering industrial alarm systems, tactical observation drones, and active decoy identification devices.

Yu said today’s cameras and surveillance systems have become faster and sharper, but still lack the information needed to identify what a target actually is. He added that SPECTRAINTEL aims to build a new sensor platform for accurately identifying materials across environmental, industrial, defense, and space applications through four-dimensional surveillance technology that captures time, space, and wavelength together — with the goal of becoming a core sensor developer for the future of physical AI.