An infrared company discovers it's only worth 10% of the business
In full-body infrared medical screening, Gaode Zhigan's (高德质感) technology provides only 10% of the diagnostic service's value, with the other 90% coming from medical expertise and field service—so the company abandoned trying to handle everything and instead shifted to selling only core components.
The video won't play here. Listen to the audio instead:
The argument · tap a timestamp to hear it
Every object emits light you cannot see
The physics of infrared is that any object with temperature emits infrared light due to molecular thermal motion; the longer the wavelength diverges from the visible spectrum, the higher its correlation with temperature. What a detector does is capture the extremely faint infrared signal from an object miles away, causing a temperature change of just fractions of a degree on the detector's surface and reading it out. It must both maintain temperature to preserve the faint signal and dissipate heat quickly to enable continuous imaging—this is the technically hardest part.
— Huang ShengFinding a lost cat was a use case the makers didn't anticipate
Infrared equipment was initially positioned for hunting and finding wildlife, but customers reported that when domestic cats go missing and hide in grass, they can't be found by calling or by eye. A sweep with infrared reveals a bright dot—get there and it's the cat. This civilian use case was not designed by the company; users discovered it themselves and fed back the insight.
— Huang ShengInfrared catches power grid failures before equipment breaks
Electrical anomalies like poor contact or short circuits first produce localized temperature spikes, and infrared inspection discovers aging components before equipment truly breaks down, causing breaker trips and blackouts—allowing engineers to plan replacement of backup equipment and avoid large-scale outages. Huang Sheng (黄胜) noted that the 1970s-80s generation experienced frequent childhood power outages, whereas now there are fewer, which relates to infrared technology's widespread adoption in power systems.
— Huang ShengWhen you provide 10% of the value, you shouldn't fight for the other 90%
In scenarios like full-body infrared medical screening, infrared is just one critical technical point in the diagnostic process; medical expertise, theoretical backing, and field service constitute ninety percent of the entire diagnostic and customer service chain. The company concluded that rather than filling gaps in capabilities it doesn't understand, it's better to do only what it does best—that ten percent—and sell equipment or chip modules to partners who truly understand medicine and Chinese medicine theory, letting them handle the remaining ninety percent. This decision determines which tracks the company runs end-to-end products (like airport body-temperature screening, where the process is simple and requires no diagnostic judgment) and which it supplies components for.
— Huang ShengCompetitors losing money does not mean the market doesn't exist
When the company evaluated potential businesses, it didn't just look at whether it would make or lose money; it also analyzed why competitors lose money—is it because the market itself has no demand, or because others haven't found the right approach? Judging whether you have an advantage to turn a money-losing business into a profitable one is the key step in deciding whether to enter a new scenario, rather than simply abandoning the entire direction because others are losing money.
— Huang ShengOnce spun off, nobody sets problems for them to solve anymore
After Gaode Zhigan (高德质感) spun off from Gaode Hongwai (高德红外) as an independent civilian company, the organization initially retained the habit of passively receiving requirements—business units would ask power grids what they needed and directly make that the product direction. But it turned out that different provincial power grids had very different needs, making it impossible to produce a single form factor satisfying all customers, with R&D costs too high. The company therefore established a dedicated product department and marketing department to synthesize scattered front-line requirements into a single product proposal acceptable to most customers, then hand it to R&D—essentially the organization had to learn to set its own problems.
— Huang ShengAI improves how images look, but doesn't command higher prices
AI applications at Gaode Zhigan fall into two categories: one is imperceptible to users, such as using machine learning to optimize image brightness adjustment in high-light and low-light scenarios, making pictures look more comfortable—but this upgrade won't make the product sell for more; it's only a user-experience improvement. The other category is recognizing the photographed object (such as transformers or insulators) and providing more appropriate algorithms and shooting recommendations, helping power-grid engineers complete inspections more efficiently—an auxiliary function rather than an independent selling point.
— Huang ShengData centers consume electricity, which means infrared has business opportunities
Infrared directly relates to energy flows: major producers like petroleum, natural gas, solar, and wind power, as well as major consumers like metallurgy, all need infrared for monitoring and maintenance. Huang Sheng (黄胜) believes that as AI compute centers become a new major energy consumer and scale up, they will also pull the infrared industry along—though not as directly as how the optical-module industry was directly boosted by AI.
— Huang ShengIn their own words · checked verbatim
You could say every object is actually emitting light—it's just that the light it emits is invisible to our naked eye.
可以理解成每一个物体其实都是在发光 只是发的这个光我们肉眼看不见
Huang Sheng4:16
The value we provide might be only 10%, and the remaining 90% might come from medical expertise.
我们提供的价值可能也就只有10% 剩下90%可能来自于这些医疗的一些认知
Huang Sheng27:44
Just because one person is losing money doesn't mean the market doesn't exist—the real question is whether you have the advantage to make that money-losing business profitable.
还是说亏钱只是他亏了 不代表这个市场不存在 我们有没有优势去把他亏钱的生意做成挣钱
Huang Sheng32:45
It won't actually sell for a higher price, but at least users will perceive the experience as more comfortable.
他其实并不会卖的更贵 但是至少用户体验上感知会更舒服
Huang Sheng37:36
How to listen
Entrepreneurs and industry observers interested in how hard-tech companies find civilian-sector applications beyond B2B, and make organizational and commercial decisions.
The opening section featuring a car ride through Wuhan and tourist recommendations (0:00-3:03) can be skipped without affecting understanding of the content that follows.