StarCarbon Suite: One-Stop Assessment and Forecasting for Carbon Market

As the world intensifies its efforts to combat climate change, carbon credits have emerged as a crucial mechanism to incentivise reductions in carbon emissions.

A*STAR I²R’s StarCarbon Suite is a comprehensive assessment and forecasting platform that integrates blockchain technology with A*STAR’s advanced AI capabilities. It allows users to perform carbon credit assessments and analyse trends in the carbon product market. Blockchain technology enhances the integrity, accessibility, and effectiveness of the carbon market by automating the process utilising smart contracts for compliance checks and green reporting.

Here’s the three main components of StarCarbon Suite.

1. Efficient Adaptive Carbon Credit Assessment

Data sources: Optical/NIR/Hyperspectral/SAR images
Analysis types: vegetation analysis, object detection/segmentation from satellite image, fast adaptive AI tech, real-time carbon monitoring.

A reliable, scalable, and cost-effective solution that accurately evaluates the carbon potential of land areas, leveraging A*STAR I²R’s unique AI-guided remote sensing foundation models to quickly localise project land cover and monitor biomass changes.

Through spatio-temporal satellite image analysis and use of unmanned aerial vehicles (UAVs), we can conduct vegetation analysis and forecast the growth conditions of target land locations. Changes in biomass are used to evaluate plant carbon sequestration, which in turn helps forecast carbon credit values. These evaluations contribute to dynamic spatiotemporal trends prediction over time.



2. Long-Term AI Forecasting for Carbon Product Price

Data sources: Historical price data, various economic indexes, monetary data, demand/supply data, carbon assessment output, news data, etc.
Analysis types: multi-step ahead price forecasting, sentimental analysis with news data, Time series LLM enhanced forecasting, hourly/daily/monthly projection, etc.

A*STAR I²R's advanced AI capabilities in time series forecasting are leveraged to predict carbon market trends. Such forecasting is essential to stabilise the market and improve investment decision-making. 

Our predictions utilise various data sources, including historical price data, demand and supply factors, economic and monetary market indexes, carbon assessment results from satellite images, and global news data. In the data analysis, functions such as web scraping, data cleaning, and correlation analysis, sentimental analysis, feature engineering and selection are performed to generate useful features for deep learning-based forecasting models. We also harness the power of prompt-based large language models to enhance forecasting accuracy. This functionality helps users make better strategic investment decisions and optimise carbon reduction strategies.

3. Blockchain for Carbon Market

• Private/Public blockchain
• Smart contract for process optimisation
• Work efficiency improvement & Cost reduction

Blockchain Technology enhances the transparency and reduces fraud. It enables efficient peer-to-peer trading of carbon credits on a global scale. The system implements blockchain solutions to improve carbon market integrity, accessibility, and overall effectiveness, fostering greater participation and accelerating progress towards global emissions reduction targets.

Carbon Market Trend Forecasting Module