At 2:21 a.m. on the 2nd (Korea time) last month, a Vega-C carrying the Korea Multi-Purpose Satellite 7 (Arirang 7) lifted off from the space center in Kourou, French Guiana. About 44 minutes after liftoff, Arirang 7 separated normally from the launch vehicle, and at 3:30 a.m., about 1 hour and 9 minutes later, it made first contact with the Troll Ground Station in Antarctica. In the process, the satellite's initial status, including the deployment of the solar panels, was also confirmed. Arirang 7 has now entered the phase where it must prove its performance in orbit.
Arirang 7 is designed to provide Earth observation imagery for public purposes such as disaster and hazard monitoring and land and environmental monitoring through a high-resolution optical camera and an infrared (IR) sensor.
Kim Sung-hee, 44, a vice president at TelePIX who took part in developing Arirang 7, met with ChosunBiz at TelePIX's headquarters in Yeouido, Seoul, on the 17th last month and said, "When people talk about CubeSats, the focus is often on building them cheaply and quickly, so there are cases where they simply send a commercial lens or camera into space without sufficient verification and analysis," adding, "Naturally, that doesn't work well. That's why there are many cases of CubeSat failures."
Drawing on the testing and verification experience accumulated during large national satellite development, Kim stressed that the smaller the satellite, the more important the procedures are for securing "verification data."
After completing a doctoral program in astronomy at Yonsei University, Vice President Kim worked for 17 years at the Korea Aerospace Research Institute (KARI) and served as a core member in developing all optical satellites launched in Korea, including the Cheollian satellite and Arirang 7. Kim handled the design, alignment, and verification of the optical system, the "eyes of the satellite." Alignment is the process of finely positioning multiple mirrors (reflectors) at the micrometer (μm, one-millionth of a meter) level to achieve focus.
After extensive experience developing optical payloads for large satellites at the institute, Vice President Kim moved to the private sector. At TelePIX, a satellite and satellite data company, Kim oversees the satellite systems institutional sector and leads key research and development (R&D), including developing electro-optical (EO) cameras for small satellites.
Founded in 2019, TelePIX touts a "space AI total solution" from hardware to artificial intelligence (AI)-based data analysis. Starting with a pre-Series A investment in 2022 and through Series B in 2025, it has reportedly raised around 30 billion won in cumulative (publicly disclosed) funding. In July 2025, it also secured investment from Japan's Elephant Design Holdings to push into the Japanese market.
◇ The minutes after launch hide time and verification
Asked about the most difficult moment on the path to launching Arirang 7, Vice President Kim recalled the initial planning rather than technical hurdles. Multi-purpose satellites have broad demand, from close-up observation to change detection. Users always want the highest performance, but the development team has to set a realistically achievable scope.
Kim explained, "Because optical equipment performance scales with size, the bigger you make it, the better it performs, but since the satellite must fit on a rocket, there is a physical limit to size," adding, "In the end, the biggest challenge in the planning phase was finding the right balance through agreement with users."
The optical system Kim handled corresponds to the lens in camera terms, but at satellite scale it is not a simple lens; it is built by combining multiple mirrors (reflectors). The more complex the structure, the higher the manufacturing difficulty, and large optical components can take 1–2 years to produce. If performance falls short of expectations or problems arise, reprocessing can cause delays of several months to a year. When production drags, assembly and alignment schedules are compressed, concentrating workload and risk in a short period during assembly and alignment.
On top of that comes the variable of differing conditions between the ground and space. An optical system built on Earth changes shape slightly in space as gravity effectively disappears. That means focus tuned under slight ground-induced sag can shift in space. Ground testing therefore requires predicting the in-orbit state and interpreting the results to determine alignment values.
Kim said the process of building a satellite is akin to an endless repetition of trials and verification. "When unplanned problems arise, we spend time finding the cause, and that experience becomes know-how that turns into test items for the next satellite," Kim explained.
So what remains from that experience? The word Kim repeated most when talking about know-how was "measurement." How to measure performance after the optical system is assembled, and how to verify under the same criteria when conditions change—gravity, atmosphere, vacuum—ultimately determines success.
As for why Arirang 7 could succeed, Kim said, "We invested heavily in building large test facilities starting 15 years ago, and we did a lot of testing while preparing for launch in advance." There are international standards worldwide to ensure reliability, such as ISO and ECSS, but most are guidelines at the level of what tests to perform. The details—how to build test facilities, what problems arise during operations, and how to overcome them—are accumulated in the field. Kim emphasized that as these details accumulate, the probability of failure decreases.
◇ Bringing the institute system to the private sector… TelePIX's "operate many" strategy
Vice President Kim is redefining the systematic development and testing procedures built at the institute to fit private development environments, establishing a tighter development protocol to reduce failure rates even for small satellites.
The direction Kim envisions at TelePIX is to launch many small, low-cost satellites to observe more frequently. "A single satellite has difficulty imaging a specific region at the same time every day, and some areas may go 2–3 days without coverage," Kim said. "Conversely, operating multiple satellites can drastically shorten revisit cycles."
To back this strategy, the company is developing a next-generation payload camera. Instead of the traditional cylindrical shape, it is closer to a rectangular form, and optically it adopts an "off-axis architecture (placing the optical axis off-center)." In a typical on-axis architecture, the light path can be blocked by as much as half as light travels back and forth, but an off-axis architecture reduces obstructions in front of the camera to boost performance.
Kim added, "The satellite equipped with this camera, 'Chouette' (French for owl), will complete development in 2026 and be launched into space for the first time in 2027."
At the same time, the company is pursuing "vertical integration" by bringing in machining personnel and performing some processes in-house to make schedule and performance management easier and, over the long term, improve the cost structure. Vice President Kim expects this approach could cut expense by up to half.
Still, Kim emphasized that "technology does not immediately become a market." "There remains the question of where this excellent technology is used. We are at a stage where demand has not yet grown in earnest," Kim said. "Only when the private sector shows, with results, why satellite imagery is necessary and changes perceptions will 'real new space' arrive."
Kim added, "When users naturally come to use satellite imagery, the model of launching many satellites to observe more frequently and analyze more will finally gain strength," adding, "Until then, the private sector's task is to build up technology and verification step by step and continue producing real use cases in the field."