Name of Research Group: Climate Seal

Description of Research Group: (1) Background and Problem Statement (Why We Did This)
Team is to address the pressing needs of the “dual carbon” strategy, green manufacturing, and the rapid development of international green trade. It tackles the key pain points faced by enterprises in carbon footprint accounting, evidence collection, compliance verification, and the recognition of green outcomes—specifically, “slow calculations, scattered evidence, difficult verification, limited reusability, and high costs.” We have developed an AI-powered self-auditing carbon accounting platform. Traditional methods rely heavily on experts manually organizing activity data, matching emission factors, interpreting standards, and repeatedly supplementing evidence. A single product carbon footprint or life cycle assessment project often takes weeks to months, incurring high costs, low efficiency, inconsistent standards, and producing results that are difficult to reuse directly by customers, financial institutions, or for regulatory purposes. This has become a significant bottleneck in the green transition for industries such as resource and energy, mining equipment, materials processing, and export-oriented manufacturing.

(2) Scientific Questions and Objectives (What We Aim to Do)
The core question this team as well as project addresses is: How can we calculate carbon emissions not only at a low cost and with high efficiency, but also ensure the results are verifiable and reusable across different scenarios? To achieve this, the project aims to solve the common problems in traditional green compliance processes: “too few qualified professionals, high verification costs, and difficulty in sharing results.” We aim to build a new technological system that can both perform carbon accounting and simultaneously generate a verifiable chain of evidence and audit records. The project has developed a technical pathway combining “AI methodological guidance + automated data processing + risk pre-verification + auditable trail logging.” On one hand, the system guides users through defining accounting boundaries, converting units, identifying data gaps, matching factors, and generating reports. On the other hand, throughout the data processing, it simultaneously conducts risk checks, links evidence, and creates traceable records, resulting in verifiable results and evidence packages. This breakthrough moves beyond the traditional sequential model of “data first, verification second,” automating and integrating carbon accounting, evidence organization, and pre-verification, enhancing the capability from single calculations to multi-scenario reuse.

(3) Research Content (How We Did It)
The research follows the overarching approach of “structured methodologies, standardized data, automated accounting, pre-emptive risk management, and process traceability.” First, we deconstructed key steps in methodologies like Product Carbon Footprint (PCF) and Life Cycle Assessment (LCA)—such as boundary definition, functional unit setting, inventory collection, factor matching, and result interpretation—into standardized rules that can be executed automatically and supported by the system. Second, to address the challenges of complex data sources, inconsistent formats, unit confusion, and frequent data gaps, we established mechanisms for data cleaning, unit conversion, field mapping, anomaly detection, and gap identification. Furthermore, we integrated a risk pre-verification module within the accounting process to perform consistency checks on key data, evidence materials, factor sources, and calculation logic. An audit trail mechanism retains a complete record of the entire process, ultimately generating green outcomes and supporting evidence that are traceable, verifiable, and reusable. The project is progressing along a path of “prototype development → pilot scenario testing → third-party validation → industry promotion,” gradually enhancing the maturity and applicability of the results.

(4) Key Achievements (What We Have Accomplished)
To date, the project has produced a prototype of the self-auditing carbon accounting and green recognition intelligence platform, successfully validated across typical scenarios including corporate carbon inventories, product carbon footprints, life cycle assessment, export compliance, and green supply chains. The platform integrates capabilities such as methodological guidance, data processing, factor matching, evidence collection, risk pre-verification, audit logging, and report generation, establishing a complete chain from raw business data to accounting results and supporting evidence. In existing pilots, tasks that previously relied heavily on manual effort—such as data compilation, modeling, evidence collection, and report generation—have seen significant workload reductions, with delivery timelines compressed from weeks or months to hours. The team has achieved LCA/PCF report generation within 2 hours, obtaining third-party verification (from BV) at a 5% materiality level. This demonstrates the practical foundation for extending this technical approach to verification-level applications. The project has also developed structured methodologies, rules, and scenario-specific delivery solutions, laying the groundwork for future technology transfer and industry replication.

(5) Application and Commercialization Value (Why It Matters)
The greatest strength of this achievement lies not just in “faster carbon calculations,” but in “making green outcomes more credible, more demonstrable, and more convertible.” Strategically, this achievement helps enhance the credibility of foundational data for China in areas like green trade, low-carbon manufacturing, green supply chains, and green finance, strengthening the capacity for autonomous responses to international regulations and customer audit requirements. At the industry level, it serves sectors such as resource and energy, mining equipment, materials processing, and export-oriented manufacturing, helping companies reduce compliance and verification costs, shorten delivery cycles, and improve efficiency in product exports and supply chain coordination. At the academic level, this achievement integrates environmental science, data science, artificial intelligence, and industrial applications. With its distinct interdisciplinary characteristics and educational value, it is well-suited for showcasing as a model outcome in the field of “AI + Green & Low-Carbon.” In the future, the platform can be further extended to serve as a green data infrastructure for mining, manufacturing, and energy-intensive industries, demonstrating significant demonstrative value and potential for widespread adoption.

Team members Introduction:

xuancheng.zhou@climateseal.net