Eco-AI: Artificial Intelligence Framework for Predicting and Reducing Carbon Emissions in Depok City

Depok City faces increasing carbon emissions associated with rapid urbanization, a population exceeding 2.7 million, and more than 821,000 registered motor vehicles. However, the city lacks an integrated, data-driven framework for predicting greenhouse gas emissions and evaluating emission reduction strategies. This study aimed to develop Eco-AI, an artificial intelligence (AI)-based framework designed to predict greenhouse gas emissions and assess mitigation strategies in Depok City through the integration of multiple predictive and optimization models. A mixed-methods approach was employed, combining quantitative and qualitative analyses using Internet of Things (IoT) sensor measurements (MQ-135 and SDS011), Sentinel-5P satellite observations, and administrative data from relevant government agencies, including DLHK, Dishub, PLN, and BPS. Five AI models were incorporated: long short-term memory (LSTM) networks for time-series forecasting, random forest (RF) for identifying influential emission-related factors, extreme gradient boosting (XGBoost) for emission hotspot classification, graph neural networks (GNNs) for spatial analysis, and reinforcement learning using proximal policy optimization (PPO) for policy scenario optimization. The results indicated that Depok City's total greenhouse gas emissions reached 4,217,340 metric tons of carbon dioxide equivalent (tCO?e) in 2024, with transportation accounting for 43.2% and residential energy consumption contributing 29.9%. The LSTM model achieved strong predictive performance (R² = 0.924; mean absolute percentage error [MAPE] = 4.97%), while the RF model identified the number of motor vehicles as the most influential predictor (feature importance = 0.234). Under the business-as-usual (BAU) scenario, annual emissions were projected to reach 5,891,200 tCO?e by 2030. The combined mitigation scenario indicated a potential emissions reduction of 19.1% relative to the BAU projection, with an estimated investment requirement of IDR 1.24 trillion. These findings suggested that the Eco-AI framework could support data-driven emissions forecasting, identification of emission-related factors, spatial assessment, and evaluation of mitigation strategies. The framework may contribute to evidence-based urban climate planning and support Depok City's efforts toward Indonesia's national net-zero emissions target for 2060.

Authors

Institutions

Publication Details

Journal
Journal Of Social Research
Published
2026-11-18
DOI
https://doi.org/10.55324/josr.v5i12.3552
Primary Topic
Air Quality Monitoring and Forecasting
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Eco-AI: Artificial Intelligence Framework for Predicting and Reducing Carbon Emissions in Depok City

Bodi Santoso, Kasih Kasih
8 links
Journal Of Social Research
Air Quality Monitoring and Forecasting
article

Eco-AI: Artificial Intelligence Framework for Predicting and Reducing Carbon Emissions in Depok City

Bodi Santoso, Kasih Kasih
article en

Abstract

Depok City faces increasing carbon emissions associated with rapid urbanization, a population exceeding 2.7 million, and more than 821,000 registered motor vehicles. However, the city lacks an integrated, data-driven framework for predicting greenhouse gas emissions and evaluating emission reduction strategies. This study aimed to develop Eco-AI, an artificial intelligence (AI)-based framework designed to predict greenhouse gas emissions and assess mitigation strategies in Depok City through the integration of multiple predictive and optimization models. A mixed-methods approach was employed, combining quantitative and qualitative analyses using Internet of Things (IoT) sensor measurements (MQ-135 and SDS011), Sentinel-5P satellite observations, and administrative data from relevant government agencies, including DLHK, Dishub, PLN, and BPS. Five AI models were incorporated: long short-term memory (LSTM) networks for time-series forecasting, random forest (RF) for identifying influential emission-related factors, extreme gradient boosting (XGBoost) for emission hotspot classification, graph neural networks (GNNs) for spatial analysis, and reinforcement learning using proximal policy optimization (PPO) for policy scenario optimization. The results indicated that Depok City's total greenhouse gas emissions reached 4,217,340 metric tons of carbon dioxide equivalent (tCO?e) in 2024, with transportation accounting for 43.2% and residential energy consumption contributing 29.9%. The LSTM model achieved strong predictive performance (R² = 0.924; mean absolute percentage error [MAPE] = 4.97%), while the RF model identified the number of motor vehicles as the most influential predictor (feature importance = 0.234). Under the business-as-usual (BAU) scenario, annual emissions were projected to reach 5,891,200 tCO?e by 2030. The combined mitigation scenario indicated a potential emissions reduction of 19.1% relative to the BAU projection, with an estimated investment requirement of IDR 1.24 trillion. These findings suggested that the Eco-AI framework could support data-driven emissions forecasting, identification of emission-related factors, spatial assessment, and evaluation of mitigation strategies. The framework may contribute to evidence-based urban climate planning and support Depok City's efforts toward Indonesia's national net-zero emissions target for 2060.

Journal Of Social ResearchVol. 5(12)
Universitas Pamulang (ID)
Industry, innovation and infrastructure, Sustainable cities and communities, Climate action
Air Quality Monitoring and Forecasting
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon