We develop advanced mathematical models and computational intelligence techniques to address critical environmental challenges β bridging theoretical innovation with real-world impact in sustainability, climate science, and ecological systems.
AMCI-Enviro is a multidisciplinary research group dedicated to advancing environmental science through the application of sophisticated mathematical and computational approaches. We believe that the complex environmental challenges facing our world require equally sophisticated analytical tools.
Our team combines expertise in applied mathematics, machine learning, optimization theory, and environmental science to develop innovative solutions for climate modeling, ecosystem management, pollution monitoring, and sustainable resource utilization. We work closely with environmental agencies, conservation organizations, and industry partners to ensure our research has meaningful real-world impact.
Through rigorous research, collaborative partnerships, and commitment to open science, we strive to contribute meaningful insights that support evidence-based environmental policy and conservation strategies.
Developing advanced mathematical models and machine learning algorithms to improve climate forecasting, analyze long-term climate trends, and assess the impact of human activities on global climate systems.
Applying computational intelligence to understand complex ecological interactions, predict biodiversity changes, and develop strategies for ecosystem conservation and restoration.
Creating intelligent systems for real-time environmental monitoring using sensor networks, satellite imagery, and advanced data analytics to track pollution, deforestation, and habitat degradation.
Designing optimization algorithms and decision support systems for sustainable resource management, renewable energy integration, and circular economy applications.
Leveraging high-performance computing and big data analytics to simulate complex ecological processes, assess conservation strategies, and predict ecosystem responses to environmental change.
Applying machine learning, statistical modeling, and data mining techniques to extract meaningful insights from large environmental datasets and support evidence-based policy making.
Sistem Prediksi Risiko Erosi Pesisir Berbasis Kecerdasan Buatan dan Simulasi Monte Carlo
Kerangka pikir penelitian ini menggambarkan alur sistematis dari pengumpulan data hingga pengembangan produk akhir AI-Coastal Risk System. Pendekatan yang digunakan mengintegrasikan multiple data sources, advanced analytics, dan machine learning untuk menghasilkan sistem prediksi yang robust dan probabilistik.
Gambar: Kerangka Pikir Riset
Tujuan Utama:
Tahapan kegiatan disusun untuk memastikan tercapainya
akhir AI-Coastal Risk System secara bertahap
selama 3 tahun, sesuai dengan peta jalan penelitian.
| Tahun | Fokus Tahapan | Komponen Utama | Luaran Antara |
|---|---|---|---|
|
2026 (Tahun 1) |
Data Foundation and Feature Extraction |
|
Dataset spasial-temporal lengkap + baseline mapping erosi |
|
2027 (Tahun 2) |
Model Development and Stochastic Simulation |
|
Model AI-Monte Carlo terverifikasi |
|
2028 (Tahun 3) |
Integration, Dissemination, and Innovation Product |
|
Produk akhir + policy guidebook |
Tahun pertama merupakan tahap dasar (foundational phase) yang berfokus pada pembentukan database spasial-temporal risiko erosi dan pra-pemrosesan citra satelit sebagai bahan utama untuk pelatihan model ML di tahun berikutnya.
Langkah ini bertujuan menyiapkan variabel input (predictor features) untuk model machine learning.
Menggunakan algoritma Normalized Difference Water Index (NDWI) dan Canny edge detection untuk mendeteksi batas darat-laut.
Perhitungan Laju Erosi: Menghitung perubahan posisi garis pantai antar tahun dengan metode End Point Rate (EPR) dan Linear Regression Rate (LRR).
Menggabungkan seluruh parameter menjadi data cube spasial-temporal (2015-2026) dengan resolusi seragam.
Dilakukan analisis statistik untuk menentukan variabel paling signifikan terhadap erosi, meliputi:
Hasil dari tahap ini:
Data multi-sumber yang terstandar untuk input model ML
Peta distribusi perubahan garis pantai historis
Laporan variabel paling berpengaruh terhadap risiko
"Coastal Change Detection in South Bali using Sentinel-2 and Machine Learning Preprocessing"
Secara keseluruhan, hasil dari tiap tahap akan berkontribusi langsung terhadap produk akhir AI-Coastal Risk System seperti pada diagram berikut:
Gambar: Integrasi menuju produk akhir
Visi Produk Akhir:
Sistem akhir akan menjadi prototipe nasional untuk mitigasi erosi berbasis data besar dan kecerdasan buatan, mendukung program Riset dan Inovasi Indonesia Maju (RIIM) serta SDGs 13 & 14 tentang aksi iklim dan perlindungan ekosistem laut.
Platform interaktif untuk visualisasi risiko erosi real-time dengan multi-layer mapping
Model ensemble (RF + XGBoost + CNN) untuk prediksi akurat dengan confidence intervals
Kuantifikasi ketidakpastian dan analisis probabilistik untuk multiple scenarios
Sistem peringatan dini dengan notifikasi otomatis untuk zona berisiko tinggi
Akses mudah melalui web dan mobile untuk stakeholder dan masyarakat
Panduan kebijakan mitigasi berbasis sains untuk pemerintah dan pengelola kawasan
Focus: Data Collection & Feature Extraction
Key Activities: Satellite imagery acquisition, preprocessing, baseline mapping, correlation analysis
Deliverables: Comprehensive dataset, baseline erosion maps, Q3 journal publication
Focus: ML Model Development & Monte Carlo Integration
Key Activities: ML training (RF, XGBoost, CNN), uncertainty quantification, probabilistic risk mapping
Deliverables: Validated AI-Monte Carlo model, probabilistic risk maps, Q1-Q2 publications
Focus: System Integration & Dissemination
Key Activities: WebGIS development, stakeholder training, policy brief creation
Deliverables: AI-Coastal Risk System (production-ready), policy guidebook, training programs
Our diverse team brings together mathematicians, computer scientists, environmental scientists, and engineers, all united by a passion for leveraging computational methods to address environmental challenges. We foster a collaborative environment that encourages interdisciplinary thinking and innovation.
Group Leader
Expert in applied mathematics and computational modeling with 15+ years of experience in environmental applications.
Machine Learning Specialist
Develops deep learning algorithms for climate prediction and environmental monitoring systems.
Optimization Expert
Specializes in mathematical optimization methods for sustainable resource management.
Computational Ecologist
Focuses on ecosystem modeling and biodiversity conservation using computational approaches.
Climate Modeling
Researching stochastic models for climate prediction and uncertainty quantification.
Environmental Data Science
Working on machine learning methods for remote sensing and land use classification.
EAJAM, 2025, Vol. 15, pp. 373β391
Pure and Applied Geophysics, 2025, Vol. 182, pp. 489β506
Jurnal Sains Teknologi, 2025, Vol. 14(1), pp. 56β66
MENDEL, 2024, pp. 33β42
MethodsX, 2023
FVCA 2023, Springer Proceedings in Mathematics & Statistics, Vol. 433
European Journal of Mechanics / B Fluids, 2023, Vol. 99, pp. 74β83
Fluids, 2021, Vol. 6(1), Article 26
Fluids, 2021, Vol. 6, Article 346
Department of Mathematics
Udayana University
Jl. Kampus Udayana, Badung, 20244, Bali, Indonesia
+62 813-3873-7730
We welcome inquiries from potential collaborators, students interested in joining our group, and organizations seeking research partnerships.