Listen to this Post
TikTok has introduced ChatTS, an innovative tool designed to interact with time series data stored in industrial historians. Licensed under MIT License, this tool aims to revolutionize IIoT (Industrial Internet of Things) and Industrial Data Analytics by enabling natural language queries for trend analysis, predictive maintenance, and anomaly detection.
You Should Know:
To leverage ChatTS for industrial automation, here are some key commands, scripts, and steps to integrate it into your workflow:
1. Setting Up ChatTS Locally
Since ChatTS is self-hosted, you’ll need GPU acceleration (A100/A800 recommended). Below are the steps to deploy it:
Clone the repository (if available on GitHub) git clone https://github.com/tiktok/chatts.git cd chatts Install dependencies (Python 3.8+) pip install -r requirements.txt Run with GPU support (CUDA required) python3 chatts_serve.py --gpu --model large-v2
2. Querying Time Series Data via CLI
ChatTS allows direct interaction with industrial historians (OSIsoft PI, Wonderware, InfluxDB). Example:
Query temperature trends from a sensor chatts-cli --query "Show me temperature anomalies in Boiler-3 last week" --source pi-server --tag "Boiler-3/Temp"
3. Automating Reports with Cron
Schedule automated reports using cron jobs:
Edit crontab crontab -e Add a daily report job 0 8 /usr/bin/python3 /path/to/chatts_report.py --output /var/reports/daily_anomalies.json
4. Docker Deployment (For Scalability)
Build the Docker image docker build -t chatts:latest . Run with GPU passthrough docker run --gpus all -p 5000:5000 chatts:latest
5. Integrating with Grafana for Visualization
Use Grafana’s API to push ChatTS insights:
curl -X POST -H "Content-Type: application/json" -d @chatts_alert.json http://grafana:3000/api/alerts
What Undercode Say:
ChatTS bridges natural language processing (NLP) and industrial time series data, making it a game-changer for predictive maintenance and OT/IT convergence. Key takeaways:
– GPU acceleration is mandatory for real-time analysis.
– MIT License allows customization for proprietary historians.
– Cron jobs and Docker ensure scalability in industrial environments.
For further exploration, check:
Expected Output:
ChatTS query result: - Boiler-3/Temp anomalies detected on 2023-10-12 (deviation: +12°C). - Recommended action: Inspect pressure valves.
References:
Reported By: Demeyerdavy Tiktok – Hackers Feeds
Extra Hub: Undercode MoN
Basic Verification: Pass ✅
