Seconds
Automated protection for critical facilities
Wire those seconds into a critical facility's automated protection or shutdown — so the alert triggers action, not just a notification.
Seismic Monitoring AI · Taiwan
Teaching AI
to read every tremor.
SeisBlue Co., Ltd. is a Taiwan-based technology company focused on seismic monitoring AI. We engineer deep-learning seismic research into systems that run reliably in production — deep-learning phase picking, near-real-time earthquake catalogs, and earthquake early warning.
About SeisBlue
From a single waveform,
to an earthquake alert.
SeisBlue's core capability is deep learning and seismology. We focus on turning continuous waveforms into trustworthy earthquake information.
We're always doing the same thing: turning elusive ground-motion signals into judgments you can act on in real time.
From signal to application
One signal,
read into three usable outcomes.
One continuous waveform comes in and unfolds along two paths — one races the clock, one sees the whole picture — landing three outcomes at three timescales, each serving a real-world need.
Continuous waveform
Raw ground-motion signal streaming in from stations, 7×24, without a break.
Seconds
Phase picking
Deep learning finds P- and S-wave arrival times amid the noise — where every judgment begins.
Seconds
Earthquake early warning
Once arrivals from a handful of stations are in, it estimates the impact — no need to wait for a full catalog — buying seconds of warning before strong shaking arrives.
Near-real-time
Earthquake catalog
Multi-station arrivals are merged to fix each event's location, magnitude, and time.
Long-term
Background & anomaly
The accumulated catalog builds background seismicity; deviations point to subsurface changes worth investigating.
In the field
Near-real-time
A foundation for monitoring & response
Give monitoring networks and disaster-response agencies a stable, reproducible near-real-time feed — a shared basis for the calls and bulletins that follow.
Long-term
Built together with you
Background and anomaly detection is where SeisBlue's deep-learning foundation comes in — we'd build it out from your data and your setting, together.
In operation
See it running.
Continuous waveforms from 472 stations across Taiwan stream in live, while the model detects and marks P-wave arrivals in step — this is the monitor during the aftershock sequence of the 2025 Dapu earthquake. The monitoring system is developed in collaboration with Taiwan's Central Weather Administration.
What We Do · Seismic AI
From research methods,
to systems that ship.
SeisBlue combines seismic-AI research experience with engineering capability — building training pipelines, integrating real-time data streams, and turning research methods into systems that run reliably in production. Below are concrete demonstrations.
Offline
Data input
Earthquake catalogs paired with continuous waveforms become labeled training data.
Offline
Preprocessing
Cleaning, windowing, and arrival-time alignment turn raw waveforms into learnable samples.
Offline
Model training
Reproducible training pipelines — experiment tracking, hyperparameters, and training strategy in one place.
Offline
Evaluation
P- and S-detection performance is validated on standard test sets — a model ships only if it beats the current version.
Online
Deployment
Packaged as a service and wired into real-time data streams, running reliably 24/7.
Online
Monitoring
Latency, performance, and data drift are watched continuously; new events picked in production accumulate as fresh training data.
↺ New events flow back as the next round of training data — the model keeps getting sharper
Core capabilities
Three stages of this pipeline —
our three core capabilities.
Method Research
From loss functions and training strategies to data engineering — deep research experience that raises performance ceilings at the model level.
Training Infrastructure
Reproducible, scalable training infrastructure — covering data preprocessing, model architecture, and training strategy.
Production Deployment
Integrating models into real-time data streams (e.g. Earthworm) to keep services running 24/7.
Method Research
BlueDisc · S-wave detection method
BlueDisc grew out of Chun-Ming Huang's years of research on phase picking. Targeting the long-standing S-wave amplitude suppression problem in deep-learning phase pickers, the work dissects the role of the loss function in training and why the data fails to converge — validated through a shape-then-align training strategy with a conditional GAN proof of concept, raising effective S-phase detections by 64% on standard benchmarks. Source at github.com/SeisBlue/BlueDisc. For a plain-language read: Why Does AI Keep Missing S-Waves? →
Huang, C.-M., Chang, L.-H., Chang, I.-H., Lee, A.-S., & Kuo-Chen, H. (2025). Recovering Sub-threshold S-wave Arrivals in Deep Learning Phase Pickers via Shape-Aware Loss. arXiv:2511.06731. arxiv.org/abs/2511.06731
Training Infrastructure
SeisBlue Platform · seismic data framework
SeisBlue is a deep-learning seismic data processing framework developed and open-sourced by Chun-Ming Huang. The platform packages the full pipeline — from waveform ingestion through phase picking to catalog output — into a service that runs reliably in production, adopted by multiple peer-reviewed studies as a core processing tool. Source at github.com/SeisBlue/SeisBlue.
Huang, C.-M., Chang, L.-H., Kuo-Chen, H., & Zhuang, Y. (2023). SeisBlue: a deep-learning data processing platform for seismology. EGU General Assembly 2023, EGU23-13927. doi.org/10.5194/egusphere-egu23-13927
Follow-up research adopting the SeisBlue Platform (4)
- Sun, W.-F., Pan, S.-Y., Huang, C.-M., Guan, Z.-K., Yen, I.-C., Ho, C.-W., Chi, T.-C., Ku, C.-S., Huang, B.-S., Fu, C.-C., & Kuo-Chen, H. (2024). Deep learning-based earthquake catalog reveals the seismogenic structures of the 2022 MW 6.9 Chihshang earthquake sequence. Terrestrial, Atmospheric and Oceanic Sciences. doi.org/10.1007/s44195-024-00063-9
- Huang, B.-S., Ku, C.-S., Lin, C.-J., Lee, S.-J., Chen, Y.-L. E., Jiang, J.-S., & Sun, W.-F. (2024). The first 30 min hidden aftershocks of the 2022 September 17, ML 6.4, Guanshan, Taiwan earthquake and its seismological implications. Terrestrial, Atmospheric and Oceanic Sciences. doi.org/10.1007/s44195-023-00059-x
- Sun, W.-F., Pan, S.-Y., Liu, Y.-H., Kuo-Chen, H., Ku, C.-S., Lin, C.-M., & Fu, C.-C. (2025). A Deep-Learning-Based Real-Time Microearthquake Monitoring System (RT-MEMS) for Taiwan. Sensors, 25(11), 3353. doi.org/10.3390/s25113353
- Kuo-Chen, H., Sun, W.-F., Kan, L.-Y., Pan, S.-Y., Yen, I.-C., Liang, S.-H., Guan, Z.-K., Liu, Y.-H., Chen, W.-S., & Brown, D. (2025). Real-Time Earthquake Monitoring with Deep Learning: A Case Study of the 2025 ML 6.4 Dapu Earthquake and Its Fault System in Southwestern Taiwan. The Seismic Record, 5(3), 320–329. geoscienceworld.org
Production Deployment
TT-SAM · real-time earthquake early warning
TT-SAM (Taiwan Transformer Shaking Alert Model) is an earthquake early warning system. Its architecture references the TEAM framework by Münchmeyer et al. (2020); Chun-Ming Huang led the AI engineering core — training environment setup, pipeline design, and production deployment. The real-time pipeline runs on the Earthworm real-time data stream.
Chen, Y.-H., Chan, C.-H., Chang, C.-C., & Ma, K.-F. (2026). A Deep Learning Framework for Peak Ground Velocity Prediction Using Multi-Station Velocity Waveforms: The Taiwan Transformer Shaking Alert Model (TT-SAM). Journal of Geophysical Research: Machine Learning and Computation. doi.org/10.1029/2025JH001005
Articles
Research & explainers
on seismic AI.
The stories behind the models, loss functions, and real-world data — written to be read.
How Did AI Learn to Pick Seismic Phases?
From every lab writing its own code to a one-line model call — PhaseNet, EQTransformer, and the SeisBench toolbox that tied the ecosystem together.
Read more →Why Does AI Keep Missing S-Waves?
The S-wave position is right, but its amplitude sinks below the detection threshold. How BlueDisc recovers it with a shape-aware loss — a 64% gain.
Read more →Contact
Work with SeisBlue.
Start with an email.
- Model integration & customization Bring seismic-AI models into your existing monitoring network or data stream.
- System integration & deployment From training pipeline to stable 7×24 production.
- Research & data engagements Phase picking, earthquake-sequence reconstruction, and method validation.
SeisBlue Co., Ltd. · 藍震有限公司
- Primaryjimmy [at] seisblue.com
- Address6F-3, No. 186, Sec. 4, Nanjing E. Rd., Songshan, Taipei
- GitHubgithub.com/SeisBlue