File size: 15,741 Bytes
1aa7fae 52992f2 8468016 52992f2 1aa7fae 52992f2 1aa7fae 52992f2 1aa7fae 52992f2 aeb81be 52992f2 aeb81be 52992f2 aeb81be 52992f2 1aa7fae 52992f2 704fdd8 52992f2 704fdd8 52992f2 704fdd8 1aa7fae 704fdd8 52992f2 1aa7fae 52992f2 1aa7fae |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 |
"""
Enhanced Streamlit application for engine predictive maintenance.
Intended to be run locally or deployed as a Hugging Face Space.
Features an interactive, modern UI with real-time predictions,
visualizations, and detailed insights.
"""
from __future__ import annotations
import os
import pandas as pd
import plotly.graph_objects as go
import plotly.express as px
import streamlit as st
import config
from inference import predict_engine_condition
def _is_running_in_hf_space() -> bool:
"""Check if app is running in Hugging Face Space."""
# HF Spaces set SPACE_ID or SYSTEM environment variable
return os.getenv("SPACE_ID") is not None or os.getenv("SYSTEM") == "spaces"
def _get_default_source() -> str:
"""Decide whether to load the model from HF or local based on env vars."""
# In HF Space, always use HF model
if _is_running_in_hf_space():
return "hf"
if config.HF_TOKEN and config.HF_MODEL_REPO:
return "hf"
return "local"
def create_gauge_chart(value: float, title: str, color: str) -> go.Figure:
"""Create a gauge chart for sensor readings."""
fig = go.Figure(go.Indicator(
mode="gauge+number",
value=value,
domain={'x': [0, 1], 'y': [0, 1]},
title={'text': title, 'font': {'size': 16}},
gauge={
'axis': {'range': [None, 100]},
'bar': {'color': color},
'steps': [
{'range': [0, 50], 'color': "lightgray"},
{'range': [50, 80], 'color': "gray"}
],
'threshold': {
'line': {'color': "red", 'width': 4},
'thickness': 0.75,
'value': 90
}
}
))
fig.update_layout(height=200, margin=dict(l=20, r=20, t=40, b=20))
return fig
def create_sensor_comparison_chart(sensor_data: dict) -> go.Figure:
"""Create a radar chart comparing sensor values."""
categories = list(sensor_data.keys())
values = list(sensor_data.values())
# Normalize values for better visualization (0-100 scale)
max_values = {
"Engine_RPM": 4000,
"Lub_Oil_Pressure": 10,
"Fuel_Pressure": 30,
"Coolant_Pressure": 10,
"Lub_Oil_Temperature": 150,
"Coolant_Temperature": 150,
}
normalized_values = [
(v / max_values.get(k, 100)) * 100 for k, v in zip(categories, values)
]
fig = go.Figure()
fig.add_trace(go.Scatterpolar(
r=normalized_values + [normalized_values[0]], # Close the loop
theta=[k.replace("_", " ") for k in categories] + [categories[0].replace("_", " ")],
fill='toself',
name='Current Readings',
line_color='#1f77b4'
))
fig.update_layout(
polar=dict(
radialaxis=dict(
visible=True,
range=[0, 100]
)),
showlegend=True,
height=400,
title="Sensor Readings Overview"
)
return fig
def main() -> None:
# MUST be first Streamlit command
st.set_page_config(
page_title="Engine Predictive Maintenance",
page_icon="π οΈ",
layout="wide",
initial_sidebar_state="expanded"
)
# Custom CSS for better styling
st.markdown("""
<style>
.main-header {
font-size: 2.5rem;
font-weight: bold;
color: #1f77b4;
text-align: center;
margin-bottom: 1rem;
}
.metric-card {
background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
padding: 1.5rem;
border-radius: 10px;
color: white;
text-align: center;
margin: 0.5rem 0;
}
.prediction-box {
padding: 2rem;
border-radius: 15px;
margin: 1.5rem 0;
box-shadow: 0 4px 6px rgba(0, 0, 0, 0.1);
}
.success-box {
background: linear-gradient(135deg, #11998e 0%, #38ef7d 100%);
color: white;
}
.warning-box {
background: linear-gradient(135deg, #ee0979 0%, #ff6a00 100%);
color: white;
}
.stSlider > div > div > div {
background-color: #1f77b4;
}
.stButton > button {
width: 100%;
background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
color: white;
font-weight: bold;
font-size: 1.2rem;
padding: 0.75rem;
border-radius: 10px;
border: none;
transition: all 0.3s;
}
.stButton > button:hover {
transform: scale(1.05);
box-shadow: 0 5px 15px rgba(0, 0, 0, 0.3);
}
</style>
""", unsafe_allow_html=True)
# Compact Header
st.markdown('<h1 style="font-size: 2rem; text-align: center; color: #1f77b4; margin-bottom: 0.5rem;">π οΈ Engine Predictive Maintenance</h1>', unsafe_allow_html=True)
st.markdown('<p style="text-align: center; color: #666; margin-bottom: 1rem; font-size: 0.9rem;">AI-Powered Engine Health Monitoring & Failure Prediction</p>', unsafe_allow_html=True)
# Sidebar
with st.sidebar:
# In HF Space, always use HF model and hide selection
is_in_space = _is_running_in_hf_space()
default_source = _get_default_source()
if is_in_space:
# In Space: completely hide model source selection, always use HF
source = "hf"
# Don't show any configuration UI in Space
else:
# Local development: show configuration section
st.header("βοΈ Configuration")
source = st.radio(
"π¦ Model Source:",
options=["local", "hf"],
index=0 if default_source == "hf" else 1,
format_func=lambda x: "π€ Hugging Face Hub" if x == "hf" else "πΎ Local File",
help="Select where to load the trained model from"
)
st.markdown("---")
st.header("π Quick Stats")
if is_in_space:
# In Space: always show HF model status
if config.HF_TOKEN and config.HF_MODEL_REPO:
st.success("β
Model Ready")
st.caption(f"Loading from: {config.HF_MODEL_REPO}")
else:
st.error("β Configuration Missing")
st.caption("Set HF_TOKEN as Space secret in Settings")
st.markdown("""
**To fix:**
1. Go to Space Settings
2. Add secret: `HF_TOKEN`
3. Restart Space
""")
else:
# Local development: check local model
if os.path.exists(config.BEST_MODEL_LOCAL_PATH):
st.success("β
Model Available")
st.caption("Trained model found locally")
else:
st.warning("β οΈ Model Not Found")
st.caption("Run training script first")
st.markdown("---")
st.header("βΉοΈ About")
st.markdown("""
This application uses machine learning to predict engine failures based on:
- Engine RPM
- Oil & Fuel Pressures
- Coolant Pressure
- Temperature Readings
**Status**: 0 = Normal | 1 = Requires Maintenance
""")
st.markdown("---")
st.caption("Built with β€οΈ using Streamlit & Scikit-learn")
# Balanced layout - inputs on left, larger visualization on right
col_input, col_viz = st.columns([1, 1.2])
with col_input:
# Input form
with st.form(key="engine_form", clear_on_submit=False):
st.markdown("### π§ Sensor Inputs")
# 2 columns for inputs
col_a, col_b = st.columns(2)
with col_a:
engine_rpm = st.number_input(
"βοΈ Engine RPM",
min_value=0.0,
max_value=4000.0,
value=800.0,
step=10.0,
help="Revolutions per minute"
)
lub_oil_pressure = st.number_input(
"π’οΈ Lub Oil Pressure",
min_value=0.0,
max_value=10.0,
value=3.0,
step=0.1,
help="bar/kPa"
)
fuel_pressure = st.number_input(
"β½ Fuel Pressure",
min_value=0.0,
max_value=30.0,
value=10.0,
step=0.1,
help="bar/kPa"
)
with col_b:
coolant_pressure = st.number_input(
"π§ Coolant Pressure",
min_value=0.0,
max_value=10.0,
value=2.0,
step=0.1,
help="bar/kPa"
)
lub_oil_temp = st.number_input(
"π‘οΈ Lub Oil Temp",
min_value=0.0,
max_value=150.0,
value=80.0,
step=0.5,
help="Β°C"
)
coolant_temp = st.number_input(
"π‘οΈ Coolant Temp",
min_value=0.0,
max_value=150.0,
value=80.0,
step=0.5,
help="Β°C"
)
submitted = st.form_submit_button("π Predict Engine Condition", use_container_width=True)
with col_viz:
st.markdown("### π Sensor Visualization")
# Real-time sensor visualization
sensor_data = {
"Engine_RPM": engine_rpm,
"Lub_Oil_Pressure": lub_oil_pressure,
"Fuel_Pressure": fuel_pressure,
"Coolant_Pressure": coolant_pressure,
"Lub_Oil_Temperature": lub_oil_temp,
"Coolant_Temperature": coolant_temp,
}
# Larger, more readable radar chart
radar_fig = create_sensor_comparison_chart(sensor_data)
radar_fig.update_layout(height=450, margin=dict(l=40, r=40, t=50, b=40)) # Larger and more readable
st.plotly_chart(radar_fig, use_container_width=True, config={'displayModeBar': False})
# Prediction results
if submitted:
inputs = {
"Engine_RPM": engine_rpm,
"Lub_Oil_Pressure": lub_oil_pressure,
"Fuel_Pressure": fuel_pressure,
"Coolant_Pressure": coolant_pressure,
"Lub_Oil_Temperature": lub_oil_temp,
"Coolant_Temperature": coolant_temp,
}
# Check if HF_TOKEN is set when using HF model
if source == "hf" and not config.HF_TOKEN:
st.error("β **HF_TOKEN not configured**")
st.markdown("""
**To fix this:**
1. Go to Space Settings β Repository secrets
2. Add secret: `HF_TOKEN` with your Hugging Face token
3. Restart the Space
Get token from: https://huggingface.co/settings/tokens
""")
st.stop()
with st.spinner("π€ Loading model and analyzing sensor data..."):
try:
result = predict_engine_condition(inputs=inputs, source=source)
except Exception as e:
st.error(
f"β **Prediction Failed**\n\n"
f"Error: {str(e)}\n\n"
f"**Troubleshooting:**\n"
f"- Ensure the model is trained: `python src/train.py`\n"
f"- Check model file exists: `models/best_model.joblib`\n"
f"- Verify HF credentials if using Hugging Face Hub"
)
return
pred_label = result["prediction"]
prob_faulty = result["probability_faulty"]
prob_normal = 1 - prob_faulty
# Compact results section
st.markdown("---")
result_col1, result_col2 = st.columns([1.5, 1])
with result_col1:
if pred_label == 1:
st.markdown(
f'<div class="prediction-box warning-box" style="padding: 1rem;">'
f'<h2 style="color: white; margin: 0; font-size: 1.5rem;">π¨ MAINTENANCE REQUIRED</h2>'
f'<p style="font-size: 1.1rem; margin: 0.5rem 0;">Engine is <b>LIKELY FAULTY</b> - Fault Probability: <b>{prob_faulty:.1%}</b></p>'
f'</div>',
unsafe_allow_html=True
)
with st.expander("π§ Recommended Actions", expanded=False):
st.markdown("""
- Schedule immediate engine inspection
- Verify sensor readings are accurate
- Review maintenance history
- Consult maintenance specialist
""")
else:
st.markdown(
f'<div class="prediction-box success-box" style="padding: 1rem;">'
f'<h2 style="color: white; margin: 0; font-size: 1.5rem;">β
ENGINE HEALTHY</h2>'
f'<p style="font-size: 1.1rem; margin: 0.5rem 0;">Engine is <b>OPERATING NORMALLY</b> - Fault Probability: <b>{prob_faulty:.1%}</b></p>'
f'</div>',
unsafe_allow_html=True
)
st.success("β
All sensors within normal ranges. Continue regular monitoring.")
with result_col2:
# Compact probability gauge
fig = go.Figure(go.Indicator(
mode="gauge+number",
value=prob_faulty * 100,
domain={'x': [0, 1], 'y': [0, 1]},
title={'text': "Fault Risk %", 'font': {'size': 16}},
gauge={
'axis': {'range': [None, 100]},
'bar': {'color': "darkred" if pred_label == 1 else "darkgreen"},
'steps': [
{'range': [0, 30], 'color': "lightgreen"},
{'range': [30, 70], 'color': "yellow"},
{'range': [70, 100], 'color': "lightcoral"}
],
'threshold': {
'line': {'color': "red", 'width': 4},
'thickness': 0.75,
'value': 70
}
}
))
fig.update_layout(height=200, margin=dict(l=10, r=10, t=30, b=10))
st.plotly_chart(fig, use_container_width=True, config={'displayModeBar': False})
# Compact metrics
col_m1, col_m2 = st.columns(2)
with col_m1:
st.metric("Normal", f"{prob_normal:.0%}")
with col_m2:
st.metric("Fault", f"{prob_faulty:.0%}")
# Compact Footer
st.markdown("---")
st.markdown("""
<div style='text-align: center; color: #666; padding: 0.5rem; font-size: 0.85rem;'>
<p>π οΈ <b>Predictive Maintenance System</b> | Built with Streamlit, Scikit-learn & Plotly | Developed by <b>Anant Tripathi</b></p>
<p style='font-size: 0.75rem; color: #888; margin-top: 0.25rem;'>β οΈ Use as decision-support tool, not replacement for expert diagnostics</p>
</div>
""", unsafe_allow_html=True)
if __name__ == "__main__":
main()
|