Emission Calculation System Using Learning Model for CO2 Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for calculating carbon dioxide emissions during product production lack accuracy and efficiency in predicting emissions for new processes based on actual values from previous jobs, leading to incomplete carbon offsetting mechanisms.
Innovation Solution
An information processing system that acquires order information, registers job information, transmits data to management apparatuses, calculates actual and predicted carbon dioxide emissions, and displays predicted emissions to terminals, utilizing a learning model to adjust hyperparameters for improved prediction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a learning model is used to predict carbon dioxide emissions for new processes, then prediction accuracy is improved, but system complexity increases
Solution Approach 1:
The patent introduces a learning model as an intermediary component between the emission calculation system and the prediction process. The learning model is trained using actual emission values from executed processes and then used to predict emissions for new processes, thereby improving prediction accuracy while isolating the complexity within the model training phase rather than the operational phase.
Solution Approach 2:
The system performs preliminary training of the learning model using historical emission data before actual prediction is needed. By pre-training the model with actual values from executed processes, the system prepares the prediction mechanism in advance, allowing accurate predictions for new processes without adding real-time computational complexity during emission calculation.
2Reliability
If actual emission values from executed processes are collected and used for training, then prediction reliability is improved, but data acquisition time and processing overhead increase
Solution Approach 1:
The system continuously collects actual emission values from executed processes and incrementally trains the learning model. Rather than requiring complete data collection before training, the system uses available data progressively, maintaining continuous improvement of prediction reliability without significant delays in data acquisition or processing.
Data Source
AI summary
A system for calculating emission, includes circuitry that acquires, from a first terminal, order information including processing conditions of one or more processes for producing a product; registers in a memory first job information based on the order information; transmits the first job information to one or more management apparatuses that manage execution of the one or more processes; acquires result information indicating a result of executing the one or more processes; calculates, based on the result information, an actual value of carbon dioxide emissions for a first job corresponding to the first job information; and in a case that second job information is registered, calculates a predicted value of carbon dioxide emissions for a second job corresponding to the second job information, based on the actual value of carbon dioxide emissions for the first job.