Problems We Want to Tackle
Renewable energy optimization
As Japan moves towards small-scale renewable energy development, it is important to carefully assess the potential of the land to ensure the success of these projects. The era of large-scale renewable energy development in Japan has come to a close, and it is now necessary to focus on maximizing the potential of smaller-scale initiatives.
Demand forecasting
Climate change has the potential to significantly impact industry supply chains. By utilizing hyper-local climate data, it is possible to predict changes in production yields and shifts in customer purchasing behavior. This information can help businesses adapt and make informed decisions.
Planning for climate change
Hyperlocal weather data can provide detailed and specific information about weather patterns and trends in a particular location. This information can be used to support planning for the effects of climate change by helping businesses, governments, and individuals anticipate and prepare for potential weather-related challenges and opportunities.
High Level Structure (Technology)
Features
Five weather measurements in high resolution
Temperature
Wind Spread/Direction
Snowfall
Precipitation
Solar Radiation
At any Location
Climate stations do not provide coverage for the entire territory, and satellite data can be low resolution. Borealis uses machine learning techniques to infer missing data at any location by analyzing weather patterns.
Past 20 Years
Currently, we have data available for the past 20 years. In the future, we will continue to add more historical data and sensors to expand the range of information we have available.
Up to 100m Scaling
Borealis' native fusion mesh has a resolution of 300 meters. With the use of artificial intelligence interpolation, mesh resolution can be further enhanced to 100 meters.
Prediction
High resolution mesh data can be used as input for forecasting algorithms. In the future, Borealis plans to also support AI-based forecasting techniques.
High Accuracy
Temperature
Average Error is ±1℃
Test Strategy
Picked 200 stations and 1 year of data for the testing
Designed multiple test sets with different distribution to evaluate the model comprehensively.
Ultra High Resolution Mesh
Classic Method
Borealis
Hourly Temperature mesh around mount Fuji and Tokyo area, calculated using data from the Amedas system by the Japan Metereological Association
To Build a Fairer,
more Sustainable Society
The world continues to change at an unprecedented speed. Our goal is to create as many opportunities as possible to make that change a positive one.
One way we are doing this is through Borealis, a data source designed to increase productivity and prevent damage from natural disasters
Borealis is one of the ways in which we are using AI technology to support businesses in their sustainability transformation efforts.
Tiago Ramalho
Co-founder and CEO
Use Cases
Solar Power Generation Due Diligence
Having the ability to visualize detailed historical data on daylight hours and snowfall conditions at a specific location will greatly speed up the process of preliminary analysis for development and make it accessible to smaller organizations.
Hazard Map for Prevention
By combining historical data on precipitation, wind speed, and disaster occurrences at the city or village level, it is possible to create a simple hazard map that can aid in the implementation of safety control measures.
Super Local Demand Forecasting
The generation of historical temperature maps with 0.1°C increments at 100-meter increments will allow for more accurate demand forecasting for individual stores or distribution centers than has previously been possible
Contact
Please send your inquiry from here.
A representative will contact you later.