Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,104 @@
|
|
| 1 |
-
---
|
| 2 |
-
license:
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
tags:
|
| 4 |
+
- oil-and-gas
|
| 5 |
+
- drilling
|
| 6 |
+
- physics-informed
|
| 7 |
+
- edge-ai
|
| 8 |
+
- realtime
|
| 9 |
+
- ml-agent
|
| 10 |
+
- deepboreai
|
| 11 |
+
library_name: deepboreai-sdk
|
| 12 |
+
pipeline_tag: tabular-classification
|
| 13 |
+
model-index:
|
| 14 |
+
- name: DeepBoreAI Agent
|
| 15 |
+
results: []
|
| 16 |
+
---
|
| 17 |
+
|
| 18 |
+
# DeepBoreAI Agent: Real-Time Predictive Drilling Model
|
| 19 |
+
|
| 20 |
+
**DeepBoreAI** delivers vendor-agnostic, physics-informed ML agents designed to predict and mitigate drilling hazards in real time. These agents are optimized for edge deployment, with live updates driven by telemetry from WITSML-compliant sources.
|
| 21 |
+
|
| 22 |
+
---
|
| 23 |
+
|
| 24 |
+
## Model Purpose
|
| 25 |
+
|
| 26 |
+
This model is part of the **DeepBoreAI ML Agent Suite** and is specialized in:
|
| 27 |
+
|
| 28 |
+
- Predicting mechanical/differential sticking
|
| 29 |
+
- Optimizing rate of penetration (ROP)
|
| 30 |
+
- Identifying hole cleaning inefficiencies
|
| 31 |
+
- Detecting washouts and mud losses
|
| 32 |
+
|
| 33 |
+
Each model is informed by a hybrid architecture that blends:
|
| 34 |
+
- Physical laws of drilling dynamics (e.g., conservation of energy, pressure balance)
|
| 35 |
+
- Online learning algorithms that adapt to new drilling conditions
|
| 36 |
+
|
| 37 |
+
---
|
| 38 |
+
|
| 39 |
+
## Use Cases
|
| 40 |
+
|
| 41 |
+
- **Real-time drilling optimization**
|
| 42 |
+
- **Anomaly detection and alerting**
|
| 43 |
+
- **Autonomous drilling guidance systems**
|
| 44 |
+
- **Rig edge computing deployments**
|
| 45 |
+
|
| 46 |
+
---
|
| 47 |
+
|
| 48 |
+
## How to Use
|
| 49 |
+
|
| 50 |
+
Install the DeepBoreAI SDK:
|
| 51 |
+
|
| 52 |
+
```bash
|
| 53 |
+
pip install deepboreai-sdk
|
| 54 |
+
```
|
| 55 |
+
|
| 56 |
+
Use this model in Python:
|
| 57 |
+
|
| 58 |
+
```python
|
| 59 |
+
from deepboreai_sdk.sdk import DeepBoreAI
|
| 60 |
+
client = DeepBoreAI()
|
| 61 |
+
|
| 62 |
+
data = {
|
| 63 |
+
"bit_depth": 2000,
|
| 64 |
+
"wobs": 15.2,
|
| 65 |
+
"rpm": 130,
|
| 66 |
+
"torque": 500,
|
| 67 |
+
"flow_rate": 400,
|
| 68 |
+
"mud_density": 1.1,
|
| 69 |
+
"annular_pressure": 80
|
| 70 |
+
}
|
| 71 |
+
|
| 72 |
+
result = client.post_telemetry(data)
|
| 73 |
+
print(result)
|
| 74 |
+
```
|
| 75 |
+
|
| 76 |
+
---
|
| 77 |
+
|
| 78 |
+
## Model Details
|
| 79 |
+
|
| 80 |
+
- **Architecture**: Physics-informed neural network with online learning
|
| 81 |
+
- **Precision**: Validated at 90%+ on historical and synthetic drilling datasets
|
| 82 |
+
- **Latency**: Optimized for <1s inference on edge devices
|
| 83 |
+
|
| 84 |
+
---
|
| 85 |
+
|
| 86 |
+
## Citation
|
| 87 |
+
|
| 88 |
+
If you use this model or DeepBoreAI, please cite:
|
| 89 |
+
|
| 90 |
+
```
|
| 91 |
+
@software{deepboreai2025,
|
| 92 |
+
author = {DeepBoreAI Team},
|
| 93 |
+
title = {DeepBoreAI: Real-Time Predictive AI Agents for Drilling},
|
| 94 |
+
year = 2025,
|
| 95 |
+
url = {https://huggingface.co/deepboreai},
|
| 96 |
+
license = {MIT}
|
| 97 |
+
}
|
| 98 |
+
```
|
| 99 |
+
|
| 100 |
+
---
|
| 101 |
+
|
| 102 |
+
## License
|
| 103 |
+
|
| 104 |
+
MIT License. Free for academic and commercial use.
|