International Conference on Technology Innovations for Crisis Management Vol.01 [ICTICM] 2026

Permanent URI for this collectionhttps://rda.sliit.lk/handle/123456789/5099

Browse

Now showing 1 - 20 of 28

Open Access
A Data-Driven Framework for Prioritizing Post-Disaster Non-Food Relief Needs: Evidence from District-Level Analysis following the ”Dithwa” disaster in Sri Lanka
(Sri Lanka Institute of Information Technology, 2026-05-21) Lokuliyana, S; Wijesiri, P; Jayakody, A; Peiris, H
Effective disaster response requires timely and data-driven allocation of relief resources. This study presents a statistical analysis of district-level and camp-level non-food relief requirements following the Dithwa disaster in Sri Lanka. A consolidated dataset comprising multiple districts and relief camps was analyzed using descriptive statistics, frequency analysis, cross-tabulation, and inferential statistical tests, including the Chi-square test and Kruskal–Wallis test. The results reveal that relief demand is highly heterogeneous across districts, with a small number of regions accounting for the majority of total requirements. Shelter and bedding items dominate the demand profile, followed by water, sanitation, and hygiene (WASH) supplies, indicating significant needs related to temporary living conditions and public health. Frequently requested items such as bed sheets, blankets, and sanitary packs suggest the feasibility of developing standardized core relief packages. However, statistically significant differences across districts highlight the necessity for adaptive, location-specific allocation strategies. The findings demonstrate the value of transforming operational disaster data into structured statistical insights to support evidence-based decision-making. The proposed approach contributes to improving resource prioritization and enhancing the efficiency of humanitarian response planning in disaster-prone regions.
Open Access
A Dynamic LUCIS Framework for Identifying Urban-Hazard Conflicts in the Kelani River Basin Using Near Real-Time Earth Observation Data
(Sri Lanka Institute of Information Technology, 2026-05-21) Aloka, O; Nayanajith, B; Dassanayake, S
Rapid urbanisation in Sri Lanka frequently bypasses formal planning, resulting in concealed land-use conflicts where infrastructure encroaches on disaster-prone areas. Traditional Land-Use Conflict Identification Strategy (LUCIS) models are based on static, outdated survey data and fail to capture realtime risks. This temporal lag leads to undetected unauthorized settlements in hazard zones increasing disaster vulnerability. This study introduces a Dynamic LUCIS framework utilizing near real-time Earth Observation (EO) data from Google Earth Engine, specifically the Dynamic World and Open Buildings datasets. Focused on the Kelani River Basin, the model identifies critical risk conflicts between urban development and flood/landslide safety zones. Validation using high-resolution drone imagery yielded an overall accuracy of 82.84%, F1-score of 0.83 and a Kappa coefficient of 0.66. The results demonstrate that EO-driven models can identify unauthorized settlements in hazard-prone areas with high precision, providing a scalable tool for resilient urban governance.
Open Access
An Intelligent Risk Aware Navigation Framework for Accident Hotspot Prediction, Real-Time Traffic Analysis, and Safety-Oriented Route Planning
(Sri Lanka Institute of Information Technology, 2026-05-21) Rathnapala, A; Seneviratne, O
Road traffic accidents remain a major public safety issue, particularly in urban regions where increasing traffic density and complex road environments contribute to higher accident risk. Existing navigation systems primarily optimize routes based on travel time or distance, without considering accident risk, which can expose drivers to unsafe road segments. This study proposes an intelligent risk-aware navigation framework that integrates accident hotspot prediction with real-time traffic analysis for safety-oriented route planning. A supervised machine learning model was trained using historical accident records obtained from Sri Lanka Police data to estimate accident risk levels across road segments. These risk scores are combined with real-time traffic information retrieved from mapping services to evaluate alternative routes based on both safety and travel efficiency. Experimental results show that the proposed model achieves an accuracy of 0.93 in predicting accident risk levels. Furthermore, the system is able to recommend routes that reduce exposure to accident-prone areas while maintaining acceptable travel time. A mobile prototype was developed to visualize accident hotspots and provide safer route commendations. The results demonstrate that integrating predictive accident analytics with real-time traffic information can significantly enhance navigation systems by enabling safety-aware decision-making and improving overall road safety
Open Access
Automated Flood Impact Detection from Remote Sensing Data using Transfer Learning: A New Zealand Case Study
(Sri Lanka Institute of Information Technology, 2026-05-21) Pathirana, N; Ranasinghe, M; Prasanna, R
Proper flood damage assessment is required for rapid emergency response and recovery planning. Highresolution drone imagery provides greater visual detail compared to satellite imagery and can support more accurate damage assessment after flood events. This work-in-progress study investigates the use of transfer learning with a pre-trained ResNet50 convolutional neural network to classify post-flood drone images. A manually annotated dataset of drone images representing damaged and non-damaged areas was used to finetune the pre-trained model and evaluate its performance using several metrics. The training process employed hyperparameter tuning to select the best model which achieved an accuracy of approximately 87%. These preliminary results demonstrate the potential of transfer learning for flood damage classification using a limited drone image dataset. Future work will focus on comparing the performance of additional deep learning models, incorporating satellite imagery and extending the approach from image-level classification to object-level damage detection.
Open Access
Automated Landslide Mapping with Open-Source Satellite Data in GEE: A Sri Lankan Case Study.
(Sri Lanka Institute of Information Technology, 2026-05-21) Lasantha H S; Prasanna R; Madusanka D A G
Rainfall-induced landslides are the most frequent natural hazards in Sri Lanka. This study developed a scalable, cloudbased landslide detection method that integrates satellite data, topographic parameters, and a machine learning framework implemented on Google Earth Engine (GEE) to support rapid, reliable landslide mapping in disaster scenarios and datascarce regions. The methodology was applied to the central highlands and the south-western parts of Sri Lanka. Multitemporal Sentinel-1 and Sentinel-2 images, along with a 12.5 m digital elevation model, were used to generate four composite datasets (M1–M4) for pixel-based classification. M1 includes spectral index differences between pre-event and post-event optical images, M2 incorporates spectral distance and spectral angle data, M3 contains topographic attributes, and M4 combines all datasets with synthetic aperture radar (SAR) amplitude ratio bands. Inter-area model training and validation were conducted using RF, CART, SVM, and GTB with independent datasets from study area-1, followed by transferability testing in a second area. Model performance was evaluated using overall accuracy, Precision, Recall, F1- Score, Kappa, and AUC. SVM-M4 performed best in interarea analysis, while GBT–M4 and CART–M3 showed supe Cloud contamination and false detections remained key challenges.
Open Access
Balancing User Engagement and Sustainable Web Design: A Comparative Study of 2D and 3D Hero Sections Using Kansei Engineering
(Sri Lanka Institute of Information Technology, 2026-05-21) Sandaruwan, R. M .T.; Somaweera, W. T. S.; Vasanthapriyan, S.; Sathyanjana, W.W.N.C.
With the increasing usage of digital platforms across the globe, the decisions made by designers in regards to web interfaces affect not only the interaction but also the sustainability of digital systems in the long-term. The primary visual entry point is the hero section which is usually the initial section of the page that the visitor looks at. However, the cost that the environment of the world would bear in case of using heavy 3-dimensional (3D) graphics in place of lighter 2-dimensional (2D) elements is rather unknown. This paper explores trade-off, applied to assess the 2D and 3D hero designs on user engagement and sustainability with the help of Kansei Engineering. A total of one hundred respondents (20-30 years old) rated 20 hero designs in ten themes. They rated clarity, immersion, excitement and rofessionalism. The PCA was used to identify the important perceptual attributes used to make decisions. Findings indicate that 3D graphics enhanced immersion and excitement with average scores of more than 4.5 out of 5 and 65 percent of the participants preferring 3D graphics. On the contrary, 2D graphics obtained a higher clarity and professionalism score, averages were 3.2 to 3.9 and were more relied upon in a situation where reliability was important. Although 3D components enhance emotional appeal, they significantly increase the computing capabilities, which is why the consideration of experience and smart design are crucial. These results provide guidelines to designers of how they can maximize user interaction and how they can facilitate sustainable web interfaces.
Open Access
Beyond Internet Dependency: LoRa for Post- Earthquake Structural Monitoring
(Sri Lanka Institute of Information Technology, 2026-05-21) Rallapudi, M.O; Vautier, A; Ngan, B; Perera, P; Prasanna, R
Acquiring structural response data during and after earthquakes is essential for assessing threats to critical infrastructure; however, conventional transmission systems rely on Internet connectivity that is often disrupted during major seismic events. This study evaluates Long Range (LoRa) radio as a low-power, infrastructure-independent alternative for transmitting building instrumentation data when conventional networks fail. Field experiments were conducted across three New Zealand cities, testing performance under varying radio configurations, elevations, and urban topographies. Results demonstrate reliable urban communication ranges of up to 1.6 km under optimal settings. Elevation was the dominant deployment factor, with packet delivery ratios (PDR) improving from 0–40% at ground level to near-100% at 500 m radial distance when transmitters were positioned on upper floors. Building-to-building links achieved 75–100% PDR with received signal strength values between −55 and −110 dBm, while terrain features such as hills and dense vegetation caused significant attenuation and occasional link failure at shorter distances. These findings confirm the feasibility of LoRa for post-earthquake structural monitoring and provide quantitative guidance for elevation-aware, terrain-conscious deployment in urban environments.
Open Access
Development of a MEMS-based Earthquake Dataset using the Raspberry Shake Network in New Zealand
(Sri Lanka Institute of Information Technology, 2026-05-21) Samiha, T.Z; Ravishan, D
Traditional seismic monitoring is often limited by the high cost of instrumentation and logistical barriers, hindering the expansion of earthquake monitoring networks especially in under-resourced regions. Low-cost MEMS-based sensors offer a scalable alternative, but require specialized datasets to train machine learning models adapted to their unique noise characteristics and sensitivity profiles. To address this, we systematically collected waveforms from approximately 4,000 earthquakes (magnitude 2.7 to the highest recorded) recorded across 89 Raspberry Shake stations in New Zealand from 2020–2025. Events were matched to nearby stations based on epicentral distance criteria (100 km for M 2.7–5.5, 150 km for M>5.5). A staged filtering pipeline using PhaseNet, EQTransformer, and GPD models, cross-validated with theoretical TauP arrivals, was applied to ensure phase pick quality across three confidence tiers. The final curated dataset comprises 918 high-confidence waveforms with validated P and S wave arrivals, alongside approximately 16,035 total waveform records spanning all quality tiers. This dataset addresses the critical scarcity of labeled training data for low-cost seismic instrumentation, enabling the development of phase pickers specifically calibrated for MEMS sensors.
Open Access
EcoSort: An Edge-Deployable Hybrid AI-IoT Framework with Decision Fusion for Automated Waste Sorting and Real-Time Bin Monitoring
(Sri Lanka Institute of Information Technology, 2026-05-21) Liyanage, V; Seneviratne, O
Abstract—Effective waste segregation remains a major challenge in urban environments because manual sorting, isolated sensor systems, and stand-alone vision models often fail to deliver the accuracy, integration, and operational visibility required for reliable deployment. This paper presents EcoSort, an edgedeployable hybrid AI-IoT framework that combines imagebased waste classification, sensor-assisted validation, decision fusion, automated sorting, and real-time bin monitoring within a single architecture. A MobileNetV3-based classifier performs lightweight visual recognition, while complementary sensor readings provide physical cues for validating ambiguous cases. The independent outputs are merged using a priority-based decision fusion layer that produces the final class label used to trigger the sorting actuator. The system is implemented as a lowcost prototype using embedded controllers, ultrasonic sensing, servo-based actuation, and a web dashboard for live fill-level visualization. In addition to end-to-end sorting, the monitoring layer generates threshold-based collection alerts when bin capacity approaches critical levels, improving operational responsiveness. The study contributes a unified design that addresses the fragmentation seen in prior waste management solutions, where classification, segregation, and monitoring are typically treated as separate subsystems. Prototype-level evaluation and implementation observations indicate that the hybrid pipeline improves classification dependability and sorting robustness compared with single-modality operation while remaining feasible for resource-constrained edge deployment. The proposed framework therefore offers a practical foundation for scalable, data-driven, and sustainable waste management in institutions and smart-city settings.
Open Access
Evaluating the Effectiveness of Virtual Reality-Based Fire Safety Training Using Performance Metrics and User Behavior Analysis
(Sri Lanka Institute of Information Technology, 2026-05-21) Nisme, N; Wijayasiri, U; Abeygunawardhana, P
Fire safety training is essential for improving emergency response; however, conventional methods often lack interactivity and fail to provide measurable performance data. Existing Virtual Reality (VR)-based training systems enhance immersion but primarily rely on subjective evaluation, limiting quantitative assessment of effectiveness. To address this limitation, a VR-based fire safety training system integrated with performance metrics and user behavior analysis was developed and evaluated. An experimental study involving 47 participants was conducted using a pre-test and post-test design. Learning outcomes, behavioral data, and user feedback were analyzed. Results indicate a significant improvement, with median scores increasing from 4 to 8 and 70.21% of participants showing improvement. The Wilcoxon signed-rank test confirmed statistical significance (p = 0.00016) with a moderate to large effect size (r = 0.525). Behavioral metrics further demonstrated measurable user performance, supporting the effectiveness of VR-based fire safety training.
Open Access
From Beach to Cliff: Adapting CoastSnap Citizen Science for Coastal Cliff Change Detection Using ML-Assisted Image Registration and Prompted Segmentation
(Sri Lanka Institute of Information Technology, 2026-05-21) aramillo-Velez, A; Gamlath, S; Chandirakumar, M; Prasanna, R; de Vilder,S; McColl, S; Tan, M.L; Stewart, C; Ambegoda, T.D
CoastSnap is a citizen science tool that uses repeat smartphone photographs from fixed stations to monitor coastal change, yet it has rarely been applied to coastal cliffs. We test a workflow for cliff change screening using a controlled pilot CoastSnap station at Ōnaero, New Zealand (iPhone 12), integrating (i) Machine Learningassisted ground control point (GCP) transfer for image registration, (ii) pinhole camera calibration and reprojection-based geo-rectification onto a curved cliff-surface model, and (iii) prompted segmentation for isolating cliff-related features. Auto-GCP benchmarking shows a tradeoff between precision and robustness: the Scale-Invariant Feature Transform (SIFT) model achieves low localisation error when successful, but is sensitive to changes in illumination. In contrast, the Local Feature ransformer (LoFTR) model provides a higher detection yield and more stable performance, but is sensitive to the threshold used. Nevertheless, it is well suited to operational use with human-inthe- loop verification. Calibration produced consistent parameters across repeat images, with focal length estimates matching device specifications. Visual segmentation based only in Regions of Interest (RoI) often merged adjacent objects, while text-guided Segment Anything Model (LangSAM) delineated cliff and debris more reliably than fracture-like features. Preliminary detection of the supply and removal of debris highlights the potential for quantifying the frequency and magnitude of mass movements at each cliff.
Open Access
How We Apply Agent-Based Modeling of Climate-Induced Internal Migration in Sri Lanka
(Sri Lanka Institute of Information Technology, 2026-05-21) Wickramanayaka, D.M.R.
Climate change has become a major driver of internal migration, especially in developing countries where agriculture-dependent populations are highly vulnerable to environmental shocks. This study presents an agent-based model (ABM) for simulating climate-induced internal migration in Sri Lanka. The model integrates seasonal climate variability, household economic conditions, crop failures, food insecurity, aid interventions, and return migration. Implemented in NetLogo, the simulation represents 300 heterogeneous household agents distributed across coastal, dry, and wet climatic zones. Results show that migration emerges through the interaction of climate stress, repeated crop failure, and economic hardship. Climate stress contributes 40% of migration pressure, crop failure 32%, and economic factors 28%. Aid interventions reduce displacement by 35% and increase return migration by 47%. The model offers a practical framework for understanding migration dynamics and evaluating climate adaptation policies in vulnerable regions.
Open Access
Insights from Cyclone Ditwah: Recommendations for Addressing Linguistic Barriers in Sri Lanka’s Disaster Early Warning Systems through Language Technology
(Sri Lanka Institute of Information Technology, 2026-05-21) Chandrasekara, S
Disaster early warning systems are critical for reducing loss of life and ensuring timely response during emergencies. However, the effectiveness of such systems depends not only on technological capacity but also on the inclusivity of communication. In linguistically diverse contexts, language barriers can significantly limit access to life saving information. This paper presents insights on the role of technology in enabling multilingual disaster early warning and communication systems in Sri Lanka. Drawing on observations from recent disaster events, specifically the catastrophic Cyclone Ditwah in 2025, and a review of existing policy and operational practices, the paper highlights gaps in language consideration within current disaster management frameworks. By analyzing local gaps, such as the incomplete localization of platforms and evaluating global models like Japan’s J-Alert and India’s SACHET, this study identifies practical, technology driven pathways for inclusion. The paper proposes the integration of AI driven translation, multilingual cell broadcasts, and text-to-speech audio alerts as core infrastructure rather than secondary features. By embedding multilingualism into disaster communication from the planning stage, authorities can ensure that urgency is communicated effectively across all communities, ultimately saving lives through clarity and trust.
Open Access
Integrated Flood, Transport, and Economic Exposure Assessment Using Sentinel-1 SAR: A Case Study of Cyclone Ditwah
(Sri Lanka Institute of Information Technology, 2026-05-21) Hettikankanama, H.S.R.; Dassanayake, S.M.; De Silva,T.S.; Mahakalanda, I.; Kaluarachchi, N.C.N.; Abeysiriwardhana,N.S.
Flooding remains one of the most destructive natural hazards worldwide, causing substantial infrastructure damage, economic disruption, and social vulnerability, particularly in rapidly urbanising tropical regions. In November 2025, Cyclone Ditwah brought intense rainfall to western Sri Lanka, leading to widespread inundation in the Gampaha Divisional Secretariat Division. This study develops and evaluates an integrated geospatial framework combining flood detection, transport network disruption analysis, and economic exposure assessment using Sentinel-1 Synthetic Aperture Radar (SAR) data. Pre-flood (1–15 November 2025) and post-flood (28 November–4 December 2025) C-band SAR imagery were processed through radiometric calibration, speckle filtering, terrain correction, and logarithmic backscatter differencing to derive a binary flood mask. The inundation layer was subsequently integrated with a connectivity-classified road network and a geocoded economic establishment dataset to evaluate transport disruption and spatial-economic exposure using an Economic Activity Exposure Index (EAEI) and Local Moran’s I clustering. The proposed workflow was evaluated using 540 stratified validation points derived from independent high-resolution reference data. Validation results yielded an overall accuracy of 81.5%, precision of 79.3%, recall of 85.2%, F1-score of 82.1%, and a Cohen’s Kappa coefficient of 0.63, indicating substantial agreement and suitability for rapid post-flood assessment. Results reveal that flooding disproportionately affected highly connected transport corridors and economically clustered zones, suggesting localised but structurally significant disruption. The study demonstrates the operational usefulness of Sentinel-1 SAR for rapid flood assessment and highlights the importance of integrating hazard mapping with infrastructure topology and economic clustering to support geospatial decision-making for disaster risk management.
Open Access
Investigating the Locational Accuracy of Crowdsourced Data in the Context of NSW Wildfires Using NLP and Geocoding Techniques
(Sri Lanka Institute of Information Technology, 2026-05-21) Vishwajith, D; Koswatte, S
Wildfires are major natural disasters, and timely monitoring is essential for effective response. Recent progress in Crowdsourced Data (CSD) and Remote Sensing (RS) has improved wildfire monitoring, however their combined use for spatial validation is still limited. This study examined the locational accuracy of Twitter-based wildfire reports during the 2019-2020 New South Wales (NSW) - Australia wildfire season by comparing them with satellite-derived hotspots. Wildfirerelated tweets were processed using Natural Language Processing (NLP), Named Entity Recognition (NER), and geocoding through the Nominatim API, while hotspot data was obtained from the VIIRS sensor via Digital Earth Australia. Advanced data cleaning and keyword filtering methods were applied to extract relevant geolocated tweets from a large raw dataset, and geocoded locations were rigorously clipped to the NSW boundary to improve spatial relevance. A 5 km grid overlay and statistical tests, including chi-square, binomial tests and Poisson-Based tests, were used to measure the spatial relationship between the two datasets. The results showed a strong positive association, indicating that Twitter reports often align with satellite-validated hotspots, particularly in densely reported areas. Although CSD can be uneven in coverage and limited by geocoding accuracy, a combined approach using NLP, spatial filtering, and statistical validation improves its reliability. The study highlighted the value of integrating social media data with RS, proposed a reproducible framework for spatial accuracy assessment, and provided practical guidance relevant to real-time wildfire monitoring and disaster management systems.
Open Access
Large Language Model Pipelines for Crisis Intelligence: Automated Help Request Classifier
(Sri Lanka Institute of Information Technology, 2026-05-21) De Silva, L.D.R.E.
This research examines the development and deployment of a Crisis Intelligence Pipeline, engineered to address the catastrophic information surge following Cyclonic Storm Ditwah in Sri Lanka [1]. The system integrates advanced prompt engineering methodologies, specifically few-shot learning, Chain-of-Thought [2], and Tree-of-Thoughts [3] to transform unstructured humanitarian data into actionable logistics strategies. Through rigorous stability testing under varying temperature gradients, the study identifies a "Safe Mode" for deterministic triage and a "Chaos Mode" that highlights the risks of model hallucination in high-stakes environments. Furthermore, the pipeline utilizes Pydantic-based schema validation [4] and token economics [5] to ensure scalability and cost-efficiency. Results indicate that structured reasoning architectures significantly enhance the precision of resource allocation and signal-to-noise discrimination, providing a vital technological framework.
Open Access
Medi-Fly-DR: A Communication-Enabled UAV Micro-Infrastructure for Disaster-Resilient Healthcare Logistics
(Sri Lanka Institute of Information Technology, 2026-05-21) Raveendran, K; Vithiyasahar, V; Karuneswaran, G; Abeygunawardhana,P.K.W
Disasters disrupt healthcare delivery by damaging transport routes and weakening the communication infrastructure required to coordinate time-critical medical logistics. This paper presents Medi-Fly-DR, a safety-aware, IoT-enabled UAV framework designed as a communication-integrated microinfrastructure for disaster-resilient healthcare logistics. Unlike conventional UAV delivery systems that primarily focus on physical transport, Medi-Fly-DR integrates autonomous medical delivery with real-time telemetry, mission-state visibility, secure data exchange, payload monitoring, dynamic geofencing, and safety-gated mission orchestration within a unified cyberphysical architecture. The framework is validated through controlled field trials using an industry-grade multirotor UAV over 1–2 km routes, supported by route-matched road baselines and fault-injection experiments. Experimental results demonstrate an approximately 80% reduction in delivery time compared with road transport, a 96% mission success rate, and safe recovery under low-battery, communication-loss, and high-wind conditions. Continuous telemetry enables real-time monitoring of mission progress, payload status, and system health, supporting traceable, coordinated, and fault-aware healthcare response. These findings show that Medi-Fly-DR functions not merely as a UAV delivery mechanism, but as an information-centric aerial infrastructure for maintaining healthcare continuity in disaster affected and resource-constrained environments.
Open Access
ResQNet: A Multi-Layer Hybrid AI Framework for Priority-Aware Disaster Communication over Decentralized Mesh Networks
(Sri Lanka Institute of Information Technology, 2026-05-21) Perera,W.M.R
Effective emergency communication is critical during disaster events, yet conventional infrastructure frequently fails through physical damage, power outages, and network congestion precisely when reliable coordination is most needed. This paper presents ResQNet, a decentralized hybrid intelligent framework designed to prioritize and route distress messages under post-disaster network conditions. The system employs a multi-layer processing pipeline integrating rule-based keyword scoring, TF-IDF with Linear Support Vector Classification (LinearSVC), a novel Distress-Aware Priority Evaluation (DAPE) heuristic, fuzzy logic-based trust evaluation, and mesh network routing using Breadth-First Search (BFS). An iterative development process upgraded the ML classifier from Multinomial Naive Bayes to LinearSVC with an expanded training corpus of 100 labeled bilingual messages, improving standalone ML accuracy from 32.5% to 70.0%. End-to-end evaluation on 80 live dashboard-tested messages demonstrates that the full hybrid pipeline achieves 71.2% classification accuracy — a 38.7 percentage-point improvement over the original ML baseline — with HIGH-priority message recall of 0.87, confirming reliable detection of safety-critical content. The system maintains stable message delivery across Dense, Sparse, and Damaged network topologies with a mean hop count of 3.4, and priority ordering effectiveness of 79.9%. Trust evaluation identified 21.2% of messages as low-credibility inputs, demonstrating resilience against noisy and misleading content. These results confirm that combining complementary AI techniques produces a significantly more robust disaster communication system than single-method approaches, with meaningful potential for real-world deployment.
Open Access
Review-Based Conceptual Framework for Generative AI in Urban Disaster Management
(Sri Lanka Institute of Information Technology, 2026-05-21) Samadhi, L. A. S. S. S; Dilshan, O.A.P; Galappaththi, K
Natural disasters, including floods and earthquakes and cyclones and wildfires, create ongoing dangers to human life and critical infrastructure and worldwide socioeconomic stability. The rising occurrence and intensity of these events require disaster management systems that can think and respond and adapt to changing needs. Generative Artificial Intelligence (GenAI) has become the dominant framework for disaster response activities. The system can create realistic simulations of disaster situations by processing extensive multimodal data. The research investigates how GenAI functions in disaster management while focusing on disaster prediction and damage assessment and response planning and resilience building. The system enables crisis management through its advanced satellite image analysis and social media text processing and sensor data evaluation and visual record assessment. The collaborative process of creating models enables urban resilience planning through scenario simulation because it shows disaster effects under different conditions. Data privacy issues together with ethical challenges and model reliability problems and misinformation risks remain as major obstacles. The study combines existing research to describe what GenAI currently achieves and what it cannot do in disaster management while presenting future research routes that will help build AI systems which are safe and reliable and beneficial for human users and will boost worldwide disaster resilience.
Open Access
Securing Distributed Disaster Early Warning Networks
(Sri Lanka Institute of Information Technology, 2026-05-21) Senarathne, K; Dias, D; Prasanna, R
This paper proposes a secure and cost-effective communication and management approach for distributed disaster early warning networks to ensure reliable, low-latency alert dissemination. While decentralized architectures reduce latency by enabling direct device communication, they are highly vulnerable to attacks due to their exposure in public community settings and the challenges posed by Network Address Translation (NAT) and firewalls. Traditional NAT traversal techniques often lack the built-in security required for such sensitive infrastructure. We identify Nebula, an open-source overlay networking tool, as a viable solution because it is lightweight enough for low-resource devices such as the Raspberry Pi typically found in decentralized early warning systems, supports NAT traversal, and provides endto- end encryption. Preliminary experiments conducted in an AWS environment—designed to simulate real-world Raspberry Shake sensor deployments—confirm Nebula’s ability to establish direct peer-to-peer (P2P) connectivity with a single hop and effectively encrypt traffic