🤝Use Case
Last updated
Last updated
SkyX has a record history of utilizing weather data for commercial use across a number of industries. Here we illustrate two sectors: power generation and disaster management.
- Commercial Application
The SkyX weather station (camera version) can be used to monitor weather and identify changes in cloud condition. It automatically records cloud information and captures all-sky cloud images in real time. This helps recognize cloud coverage and movement. Then based on our all-sky image data and algorithm products, ultra-short-term optical power forecasting can stabilize photovoltaic power generation and help with power prediction and trading.
- Performance Characteristics
Our algorithm is a mathematical model based on a combination of neural network structure and hyperlocal cloud image. Having been trained properly, the model can accurately approximate non-linear mapping. Under variable weather conditions such as cloudiness, the solar irradiance exhibits irregular and complex changes influenced by various meteorological and hyperlocal characteristic factors, in this case the traditional model will not be relevant. By incorporating the hyperlocal cloud image and meteorological data while reducing the dimensions of input variables, our model can effectively extract the key information due to weather changes, and make better quality short-term irradiance forecast, improving accuracy level from 40% to more than 90%.
Flood disaster caused by heavy rainfall during flood period is a huge threat to people’s lives and properties. There is strong demand in flood prevention management including pre-disaster preparation, emergency response and post-disaster recovery. This requires real-time control and forecast of potential risk factors such as weather, geology, water level etc. SkyX’s data network and related services can deal with such challenges.
- Torrential Flooding Management in Rural Areas
The Emergency Authority makes overall arrangements for flood prevention for the villages near a populated city. However, the relevant authority historically depends on manpower and experience to make decisions. In addition, they lack meteorological monitoring and hence some key data are absent such as precipitation. This results in difficulty in flood control and emergency evacuation. We installed devices effectively collect on-site meteorological and hydrological information, combined with public weather data, our flood perception system can timely alert the sudden rainfall and abnormal water levels. After the service was launched, the local authority can now master the flood situation and assess risk level in real time. As a result, there was no casualties caused by asymmetric flood information since 2023.
- Smart City & Flooding Management in City Areas
One major city with more than 20 million people, historically there was a lack of systematic approach to assess the risk of flood and waterlogging within the city. Local authorities were eager to seek digital solutions to manage the process. As a service provider, we built a weather data network which effectively cover the blind spots such as water-prone points, low-lying courtyards, overpasses and underground passages. We also developed flood risk management data system and standardized the decision-making process. As a result, all information and data are aggregated on one platform, innovative risk prediction and alert are embedded, the emergency and flood risk assessment system is constructed. Based on our data and forecast, local authorities are now capable of timely identifying risk spot, making informed decisions and acting accordingly.
- Flooding Management in Torrent Ditches and Small & Medium-Sized Watersheds
Many small watersheds and torrent ditch are in the suburbs. Continuous or short-term heavy rainfall can lead to flash floods, mudslides, landslides and even small barrier lake. This poses a great threat to populated areas and facilities along the middle and lower reaches of the river basin. There was no digital monitoring means for this. By building a weather data network, we digitally restore the hyperlocal data. This is combined with geographic data, elevation data, and then incorporated into meteorological models to predict precipitation, water level changes, and flash flood risk. Our network helps customers to timely grasp the on-site information, perceive sudden rainfall and abnormal water levels and make effective decisions. At the same time, when flood is formed upstream, alarm will be activated along the middle stream and downstream, avoiding casualties caused by flash floods.
