Hong Kong-based climate technology company Stellerus Tech is showcasing satellite remote sensing, Physical AI and environmental intelligence systems at Eco Expo Asia 2026, held October 26–29 at AsiaWorld-Expo in Hong Kong. Its demonstrations span extreme weather forecasting, urban flood prediction, energy infrastructure protection, marine monitoring and landslide risk assessment. The company is also highlighting two space-based initiatives: the MUSICO greenhouse gas monitoring instrument aboard China’s Tiangong Space Station and its planned Feilian hyperspectral satellite constellation.
Stellerus Connects Satellite Data With Physical AI
Stellerus Tech’s presence at the 21st Eco Expo Asia reflects a growing role for AI and satellite-based Earth observation in climate technology. The Hong Kong company, incubated by the Hong Kong University of Science and Technology (HKUST), is positioning its platform around the integration of satellite, airborne and ground-based observations with physics-guided AI models.
The commercial objective is to turn environmental data into operational intelligence for governments, infrastructure operators and businesses. Applications include forecasting severe weather, assessing flood exposure, monitoring greenhouse gas emissions and evaluating risks to energy assets.
The wider market is moving in this direction as climate-related disruptions create demand for more localized forecasting and risk information. Eco Expo Asia itself brought together more than 340 exhibitors from 15 countries and regions in 2026, providing a platform for companies working across green energy, environmental protection and climate-related technologies. ecoexpoasia.
MUSICO Brings Greenhouse Gas Monitoring Into Orbit
One of the central research initiatives highlighted by Stellerus is MUSICO, or the Multi-Spectral Imaging Carbon Observatory. Developed by an HKUST research team, the instrument arrived at China’s Tiangong Space Station aboard the Tianzhou-10 cargo spacecraft on May 11, 2026. HKUST describes it as Hong Kong’s first scientific payload deployed on the national space station.
MUSICO is designed to detect carbon dioxide and methane emissions from space at approximately 100-meter spatial resolution. Its imaging technology analyzes how sunlight is absorbed across specific spectral bands as it passes through the atmosphere and reflects from Earth’s surface. The resulting measurements can help identify individual emission sources, including industrial facilities and landfills.
This level of observation could support corporate carbon accounting, emissions verification and methane leak identification. Satellite-based measurements offer a way to observe emissions over wide geographic areas, complementing ground-based monitoring and company-reported data.
Stellerus is a sponsor of the MUSICO project and says it is working to commercialize space-based observation data for corporate carbon management, emissions verification and green finance. The commercial opportunity will depend on measurement accuracy, data availability, validation methods and the ability to translate satellite observations into information that businesses and regulators can use.
Feilian Constellation Targets 3D Wind Intelligence
Stellerus is also advancing the Feilian Constellation, a planned hyperspectral satellite system designed to observe atmospheric water vapor movements and derive three-dimensional wind field data.
The company describes Feilian as the world’s first hyperspectral satellite constellation designed specifically for 3D wind field observations. Its stated targets include 25 times higher global wind field resolution and more than 20% improvement in forecast accuracy. These are company-stated ambitions rather than independently verified operational results.
The underlying approach uses hyperspectral infrared observations to track atmospheric features, including water vapor, and apply algorithms to reconstruct wind fields. HKUST’s technology transfer office describes Project Feilian as a paired-satellite concept intended to deliver global 3D wind data at kilometer-scale resolution with an hourly revisit rate. atw.hkust.edu.hk+1
More detailed atmospheric wind observations could benefit weather prediction, renewable energy forecasting and aviation safety. Wind conditions influence turbine output and grid planning, while more accurate information about atmospheric movement can improve the modelling of severe weather systems.
Stellerus plans to combine Feilian data with proprietary hyperspectral payloads and Physical AI algorithms to develop next-generation weather models. Its longer-term goal is an Earth intelligence platform that integrates satellite observations, meteorological models and AI to provide decision support for extreme weather, disaster warnings and infrastructure operations.
Five Applications Connect Climate Data With Urban Operations
At Eco Expo Asia, Stellerus is demonstrating five applications that translate environmental observations into risk information for urban and industrial users.
– Typhoon wind field monitoring: The Hong Kong Coastal Typhoon Wind Field System provides historical typhoon replays, wind-speed queries and analysis at different altitudes. These functions are intended to support emergency planning and assessments of wind exposure.
– 3D urban flood digital twin: The system combines terrain information, drainage networks and 3D simulations to estimate flood-prone areas and water depths during heavy rainfall. Such information can help authorities identify road risks and coordinate emergency responses.
– Energy infrastructure resilience: By combining typhoon tracks, wind forecasts and power-grid information, the platform supports decisions involving wind turbine shutdowns, substation flood exposure and power-load management.
– Marine environmental monitoring: Three-dimensional ocean current simulations provide information on current speed, direction and changes, supporting coastal flood prevention, storm surge assessment and marine operations.
– Satellite-based slope monitoring: InSAR satellite observations, optical imagery and ground-based GNSS sensors are combined to track slope movement and support landslide early-warning systems.
The applications illustrate how climate intelligence can extend beyond forecasting. By connecting environmental observations with infrastructure data and digital models, operators can use forecasts to evaluate specific assets, locations and potential disruptions.
Physical AI Moves From Research Toward Climate Services
Stellerus uses the term Physical AI to describe its approach to combining physical observations, scientific models and AI algorithms. Unlike AI applications that primarily process text or business records, this approach is intended to interpret conditions in the physical environment and support decisions involving weather, terrain, water and infrastructure.
The company is not alone in exploring this direction. Research institutions and technology providers are increasingly combining satellite imagery, numerical models and machine learning to improve environmental monitoring and forecasting. Stellerus’ distinction is its focus on connecting proprietary satellite data with a range of climate-risk products.
Its challenge now is to turn research and demonstrations into dependable services. That requires validating model performance, maintaining consistent data coverage and showing that forecasts and risk assessments improve operational decisions. Commercial adoption will also depend on partnerships with governments, utilities, insurers and infrastructure operators.
The company’s Eco Expo Asia demonstrations show how it intends to connect space-based sensing with practical climate applications. If its satellite initiatives and Physical AI models deliver the targeted improvements, Stellerus could strengthen its position in the emerging market for climate intelligence and infrastructure resilience.
Market Landscape
AI-driven climate technology is developing around the convergence of satellite remote sensing, numerical weather prediction, digital twins and machine learning. Organizations increasingly need localized information about floods, storms, landslides, emissions and energy infrastructure exposure. Stellerus is targeting this demand with an integrated Earth intelligence approach, combining space-based observations with physical models and AI decision-support applications. Its commercial progress will depend on data quality, forecasting validation, deployment partnerships and the ability to demonstrate measurable value for climate-risk management.
Top Insights
- Stellerus Tech showcased five climate applications spanning typhoon monitoring, urban flood prediction, energy infrastructure, marine environments and landslide risk.
- MUSICO, an HKUST-developed instrument aboard Tiangong, monitors carbon dioxide and methane emissions at approximately 100-meter spatial resolution.
- The planned Feilian Constellation aims to provide global three-dimensional wind field observations using hyperspectral satellite technology.
- Stellerus is combining satellite observations, meteorological models and Physical AI to develop an integrated Earth intelligence platform.
- Commercial adoption will depend on validated forecasting performance, reliable satellite data and partnerships with infrastructure operators and public agencies.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
