Abstracts Submission
1. Abstract Submission Guidelines
Authors are invited to submit abstracts reflecting original, high-quality research within the scope of IconMET 2026. Adherence to the following guidelines is mandatory for consideration.
Formatting and Length Requirements
- Word Count: Abstracts must be between 200 and 250 words, excluding title, authors, and keywords.
- Font: Times New Roman or Arial, 12pt, single-spaced.
- Structure: The abstract must clearly state the Motivation/Background, Methodology, Key Results, and Conclusion/Impact.
- Keywords: Include 4 to 6 relevant keywords separated by semicolons.
- File Format: Submissions must be in Microsoft Word (.docx) format, strictly following the official IconMET 2026 template structure (Section 2).
Review and Selection Process
All submitted abstracts will undergo a rigorous double-blind peer-review process. Acceptance decisions will be based on scientific merit, novelty, relevance to the conference themes, and overall clarity.
Key Dates
Milestone | Date (UTC) |
Abstract Submission Deadline | 2026-05-18 23:59 |
Notification of Acceptance | 2026-05-25 |
Early Registration Deadline | 2026-06-01 |
Final Manuscript Submission (for accepted papers) | 2026-06-01 |
Note: Late submissions will not be accepted under any circumstances.
2. IconMET 2026 Abstract Template Structure
Authors must structure their abstract content according to the mandatory fields outlined below. Use the provided Word template file for final submission, but this outline details the required textual components.
Title (Max 15 words)A concise, informative title reflecting the core contribution.Author(s) and AffiliationsFull names, primary affiliation, and corresponding author email marked with an asterisk (*).Abstract Body (200-250 Words)Keywords4 to 6 specific terms for indexing. |
3. IconMET 2026 Abstract Sample (Illustrative Content)
Optimizing Distributed Sensor Networks for Real-Time Structural Health Monitoring using Federated LearningDr. Alistair Finch*, Dr. Ben Carter, Prof. Clara Diaz. Department of Civil Informatics, Metropolis University.
The methodology employs a hybrid CNN-LSTM model trained locally on time-series acceleration data harvested from bridge monitoring systems. We investigate the convergence efficiency of three global aggregation algorithms—FedAvg, FedProx, and Scaffold—under conditions simulating intermittent sensor connectivity (drop rates up to 20%). The key results demonstrate that the FedProx algorithm achieves model accuracy parity (within 1.5% deviation) compared to a centralized baseline, but reduces required communication overhead by 65% across the simulation duration. Furthermore, the model exhibits superior robustness against data drift caused by environmental temperature fluctuations, identified via sensitivity analysis. The principal contribution is the validation of an efficient, privacy-preserving framework for predictive maintenance, significantly lowering the operational cost barrier for large-scale SHM deployment. This work bridges advanced distributed machine learning with resilient civil engineering practice. Keywords: Structural Health Monitoring; Federated Learning; Edge Computing; CNN-LSTM; Vibration Analysis; Data Privacy. |