Smart Campus Intelligence System — Complete README

Version: 1.0
Date: April 2026
Status: ✅ Production Ready

System Overview

The Smart Campus Intelligence System is a comprehensive data analytics platform for monitoring and optimizing campus resource utilization. It processes data from 4 key sources (electricity, WiFi, mess services, attendance) and delivers actionable insights through an interactive multi-page dashboard.

Key Features

✅ 6-Module Pipeline — Modular design from data generation to dashboard
✅ 4 Data Sources — Electricity, WiFi, Mess services, Student attendance
✅ 10 Dashboard Tables — Optimized fact and dimension tables
✅ 6 Interactive Pages — Overview, utilization, peak hours, insights, anomalies, recommendations
✅ 15+ KPI Measures — Pre-built DAX formulas for instant analytics
✅ Smart Recommendations — ML-based insights with ROI calculations
✅ Anomaly Detection — Real-time outlier identification
✅ Multi-Slicer Support — Date, building, location, time-based filtering

System Architecture

Module Hierarchy

┌─────────────────────────────────────────────────────────────┐
│ MODULE 1: Data Generation (Optional — uses sample data)    │
│ Output: data/raw/*.csv                                      │
└────────────────────────────┬────────────────────────────────┘
                             ↓
┌─────────────────────────────────────────────────────────────┐
│ MODULE 2: Preprocessing & Cleaning (REQUIRED)              │
│ • Timestamp parsing, missing value handling                │
│ • Date/time feature extraction (hour, day, week)           │
│ • Anomaly flagging, crowd ratio calculation                │
│ Output: data/cleaned/*.csv                                  │
└────────────────────────────┬────────────────────────────────┘
                             ↓
┌─────────────────────────────────────────────────────────────┐
│ MODULE 3: Analytics & Aggregation (REQUIRED)               │
│ • Peak hour identification, daily/weekly trends             │
│ • Building/location utilization analysis                    │
│ Output: outputs/*.csv (peak_hours, daily_trends, etc.)      │
└────────────────────────────┬────────────────────────────────┘
                             ↓
┌─────────────────────────────────────────────────────────────┐
│ MODULE 4: ML Models (Optional)                             │
│ • Predictive electricity/mess forecasting                   │
│ Output: models/*.pkl, outputs/predicted_*.csv               │
└────────────────────────────┬────────────────────────────────┘
                             ↓
┌─────────────────────────────────────────────────────────────┐
│ MODULE 5: Recommendations (REQUIRED for Page 6)             │
│ • Smart recommendations with priority & ROI                │
│ Output: outputs/recommendations.csv                         │
└────────────────────────────┬────────────────────────────────┘
                             ↓
┌─────────────────────────────────────────────────────────────┐
│ MODULE 6: Dashboard Preparation (REQUIRED)                 │
│ • Fact/dimension table generation                           │
│ • Calculated columns for BI tools                          │
│ Output: outputs/dashboard/*.csv (10 files)                  │
└────────────────────────────┬────────────────────────────────┘
                             ↓
┌─────────────────────────────────────────────────────────────┐
│ LOCAL WEBSITE DASHBOARD                                     │
│ • 6 interactive pages with 30+ visuals                      │
│ • Dynamic cross-filtering & drill-down                      │
│ Output: local Flask web app at http://127.0.0.1:5000        │
└─────────────────────────────────────────────────────────────┘

Quick Start (5 Minutes)

Step 1: Run Data Pipeline

# Prerequisites: Python 3.8+, pandas, numpy, scikit-learn (optional)
pip install -r requirements.txt

# Run ALL modules in sequence (recommended)
python module2_preprocessing.py
python module3_analytics.py
python module5_recommendations.py
python module6_dashboard_prep.py

Step 2: Launch the Local Website

Run the dashboard app:

python app.py

Open http://127.0.0.1:5000 in your browser
Refresh the page after rerunning the pipeline
Review energy, WiFi, mess, attendance, and recommendation insights

Step 3: Explore and Share

Use the website to interact with the dashboard
Re-run the pipeline when new data arrives
Refresh the page to update the charts
Share the local URL with stakeholders on the same network

File Structure

Campus_Intelligence_System/
├── README.md                          ← You are here
├── DASHBOARD_GUIDE.md                 ← Complete BI setup guide
├── QUICK_REFERENCE.md                 ← Printable reference card
├── requirements.txt                   ← Python dependencies
│
├── module1_data_generation.py          (Optional — creates raw data)
├── module2_preprocessing.py            (REQUIRED)
├── module3_analytics.py                (REQUIRED)
├── module4_ml_models.py                (Optional — ML predictions)
├── module5_recommendations.py          (REQUIRED)
├── module6_dashboard_prep.py           (REQUIRED — THIS IS KEY)
│
├── data/
│   ├── raw/                           (Sample data if using module1)
│   │   ├── attendance.csv
│   │   ├── electricity.csv
│   │   ├── mess_entry.csv
│   │   └── wifi_usage.csv
│   └── cleaned/                       (Module 2 output — clean data)
│       ├── attendance_clean.csv
│       ├── electricity_clean.csv
│       ├── mess_clean.csv
│       └── wifi_clean.csv
│
├── models/                            (Module 4 output — ML models)
│   ├── electricity_predictor.pkl
│   ├── mess_predictor.pkl
│   └── scaler.pkl
│
├── outputs/                           (Intermediate & final outputs)
│   ├── peak_hours.csv                 (Module 3)
│   ├── daily_trends.csv               (Module 3)
│   ├── weekly_trends.csv              (Module 3)
│   ├── electricity_utilization.csv    (Module 3)
│   ├── wifi_utilization.csv           (Module 3)
│   ├── detected_anomalies.csv         (Module 3)
│   ├── mess_peak_hours.csv            (Module 3)
│   ├── recommendations.csv            (Module 5)
│   ├── recommendations.txt            (Module 5 — human readable)
│   ├── predicted_electricity.csv      (Module 4 — optional)
│   ├── predicted_mess.csv             (Module 4 — optional)
│   │
│   └── dashboard/                     (Module 6 output — READY FOR BI)
│       ├── fact_electricity.csv       ✓ 4,455 rows
│       ├── fact_wifi.csv              ✓ 4,458 rows
│       ├── fact_mess.csv              ✓ 1,302 rows
│       ├── fact_attendance.csv        ✓ 16,745 rows
│       ├── fact_peak_hours.csv        ✓ 24 rows
│       ├── fact_daily_trends.csv      ✓ 31 rows
│       ├── fact_weekly_trends.csv     ✓ 5 rows
│       ├── fact_elec_utilization.csv  ✓ 6 rows
│       ├── fact_wifi_utilization.csv  ✓ 6 rows
│       └── dim_recommendations.csv    ✓ 27 rows

Module Details

Module 2: Data Preprocessing & Cleaning

Purpose: Transform raw data into analysis-ready format
Input: data/raw/*.csv (sample data)
Output: data/cleaned/*.csv
Key Operations:

Timestamp parsing & validation
Missing value handling (forward fill, interpolation)
Feature extraction (hour, day_of_week, week, month, is_weekend)
Outlier flagging via IQR method
Crowd ratio calculation (mess)
Data quality validation

Run:

python module2_preprocessing.py

Module 3: Analytics & Aggregation

Purpose: Calculate metrics & identify patterns
Input: data/cleaned/*.csv
Output: outputs/*.csv (peak_hours, trends, etc.)
Key Calculations:

Peak hour identification (top 25% users/energy)
Daily & weekly aggregations
Building/location utilization metrics
Anomaly summary statistics

Run:

python module3_analytics.py

Module 4: ML Models (Optional)

Purpose: Predict future resource usage
Input: data/cleaned/*.csv
Output: models/*.pkl, outputs/predicted_*.csv
Models:

ARIMA + Linear Regression (Electricity)
Seasonal Decomposition (Mess)

Run:

python module4_ml_models.py

Module 5: Recommendations

Purpose: Generate smart, actionable recommendations
Input: outputs/*.csv, data/cleaned/*.csv
Output: outputs/recommendations.csv
Recommendation Categories:

Energy Saving (High ROI)
Mess Timing (Medium ROI)
Class Scheduling (Low ROI)
WiFi Optimization (Medium ROI)

Run:

python module5_recommendations.py

Module 6: Dashboard Preparation ⭐ THIS IS THE KEY

Purpose: Generate dashboard-ready tables
Input: data/cleaned/*.csv, outputs/*.csv
Output: outputs/dashboard/*.csv (10 files)
Key Features:

Automatic calculated columns (hour_label, day_name, peak_indicator, etc.)
Dimension table with auto-increment IDs
Estimated savings extraction
FutureWarnings eliminated
Comprehensive error handling

Run:

python module6_dashboard_prep.py

Expected Output:

======================================================================
  ✓ SUCCESS: All dashboard CSVs saved to outputs/dashboard/

  ✓ fact_electricity: 4455 rows
  ✓ fact_wifi: 4458 rows
  ✓ fact_mess: 1302 rows
  ✓ fact_attendance: 16745 rows
  ✓ fact_peak_hours: 24 rows
  ✓ fact_weekly_trends: 5 rows
  ✓ dim_recommendations: 27 rows
  ✓ fact_daily_trends: 31 rows
  ✓ fact_elec_utilization: 6 rows
  ✓ fact_wifi_utilization: 6 rows
======================================================================

Dashboard Overview

6 Interactive Pages

📊 PAGE 1: Overview

Purpose: Executive summary of campus utilization
Visuals: 4 KPI cards + daily electricity trend
Slicers: Date range
KPIs: Avg electricity, peak mess hour, max WiFi users, avg attendance

🏗️ PAGE 2: Resource Utilization

Purpose: Building & location-level resource consumption
Visuals: 2 bar charts (by building & location) + energy pie chart
Slicers: Building, Location
Metrics: Total electricity, WiFi users, energy distribution

⏰ PAGE 3: Peak Hours Analysis

Purpose: Identify capacity bottlenecks & peak periods
Visuals: Heatmap (WiFi hour×day) + 2 line charts (electricity, WiFi)
Slicers: Hour, Week type
Metrics: Average users/units per hour, peak indicators

📈 PAGE 4: Predictive Insights

Purpose: Trend analysis & forecasting (if ML models available)
Visuals: Multi-line weekly trends + area chart (electricity)
Slicers: Week range
Metrics: Weekly WiFi, mess entries, attendance, electricity
Note: Shows "Actual vs Predicted" if module4 was executed

🚨 PAGE 5: Anomaly Detection

Purpose: Flag unusual electricity consumption patterns
Visuals: Scatter plot + anomaly records table + count card
Slicers: Anomaly flag, Date range
Metrics: Anomaly events, affected buildings, consumption spikes

💡 PAGE 6: Recommendations

Purpose: Display AI-generated optimization recommendations
Visuals: Dynamic table with priority-based filtering
Slicers: Category, Priority
Metrics: 27+ recommendations with ROI, priority ranking, categories

Power BI Implementation Quick Reference

Create Relationships (Model View)

drag:                           to:
─────────────────────────────────────────────────────
fact_daily_trends[date]     → fact_electricity[date]
fact_daily_trends[date]     → fact_wifi[date]
fact_daily_trends[date]     → fact_mess[date]
fact_daily_trends[date]     → fact_attendance[date]
fact_peak_hours[hour]       → fact_electricity[hour]
fact_peak_hours[hour]       → fact_wifi[hour]
fact_weekly_trends[week]    → fact_daily_trends[week]

Paste Essential Measures (New Measure)

Avg Electricity (kWh) = AVERAGE(fact_electricity[units_consumed])

Max WiFi Users = MAX(fact_wifi[number_of_users])

Energy Share % = DIVIDE(
    SUM(fact_electricity[units_consumed]),
    CALCULATE(SUM(fact_electricity[units_consumed]), ALL(fact_electricity))
) * 100

Anomaly Count = CALCULATE(
    COUNT(fact_electricity[is_anomaly]),
    fact_electricity[is_anomaly] = 1
)

See DASHBOARD_GUIDE.md for complete DAX formulas (15+ measures)

Global Slicers (Enable Cross-Page)

View → Sync Slicers → check all pages for:

Date Range → fact_daily_trends[date]
Building → fact_electricity[building]
Location → fact_wifi[location]
Week Type → fact_electricity[week_type]
Hour → fact_peak_hours[hour_label]

Tableau Implementation Quick Reference

Create Relationships (Data Source Tab)

Drag fact_daily_trends to canvas, then add:

fact_electricity → relate on [date]=[date]
fact_wifi → relate on [date]=[date]
fact_mess → relate on [date]=[date]
fact_attendance → relate on [date]=[date]
fact_peak_hours → relate on [hour]=[hour]

Create Calculated Fields

Hour Label

STR(INT([Hour])) + ":00–" + STR(INT([Hour])+1) + ":00"

Peak Indicator

IF [Units Consumed] > 20 THEN "High"
ELSEIF [Units Consumed] > 10 THEN "Medium"
ELSE "Low" END

Attendance Rate %

AVG([Is Present]) * 100

Energy Share %

SUM([Units Consumed]) / TOTAL(SUM([Units Consumed])) * 100

See DASHBOARD_GUIDE.md for complete calculated field formulas (8+ fields)

Color Palette (Use Consistently)

Status Colors

🔴 High/Alert/Peak: #E74C3C (Red)
🟠 Medium/Warning: #E67E22 (Orange)
🟢 Low/Safe/Normal: #2ECC71 (Green)
🔵 Info/WiFi: #3498DB (Blue)

Building Colors (Fixed)

Academic Block A: #3498DB
Academic Block B: #2980B9
Hostel Block 1: #E74C3C
Hostel Block 2: #C0392B
Library: #9B59B6
Sports Complex: #1ABC9C

Priority Colors (Recommendations Table)

High: bg #FADBD8 | text #E74C3C
Medium: bg #FDEBD0 | text #E67E22
Low: bg #D5F5E3 | text #27AE60

Data Dictionary

Common Fields Across Tables

Field	Type	Range	Description
`timestamp`	DateTime	2024-03-01 onwards	Full datetime of event
`date`	Date	2024-03-01 onwards	Date only (no time)
`hour`	Integer	0-23	Hour of day (24-hour format)
`day_of_week`	Integer	0-6	Monday (0) to Sunday (6)
`week`	Integer	1-52	ISO week number of year
`is_weekend`	Integer	0 or 1	0=Weekday, 1=Weekend
`is_anomaly`	Integer	0 or 1	0=Normal, 1=Outlier detected
`building`	Text	"Academic Block A", etc.	Building name
`location`	Text	"Academic Zone", etc.	WiFi location name
`meal_type`	Text	"Breakfast", "Lunch", etc.	Meal service type

Calculated Columns (Added by Module 6)

Column	Calculation	Example
`hour_label`	str(hour) + ":00–" + str(hour+1) + ":00"	"09:00–10:00"
`day_name`	Map day_of_week to name	"Friday"
`week_type`	"Weekend" if is_weekend else "Weekday"	"Weekday"
`week_label`	"Week " + str(week)	"Week 9"
`month_name`	Format date as "Mon YYYY"	"Mar 2024"
`peak_indicator`	"High" if >20, "Medium" if 10-20, "Low"	"High"
`load_level`	"High" if >150, "Medium" if 50-150, "Low"	"Medium"
`crowd_level`	"High" if >0.6, "Medium" if 0.3-0.6, "Low"	"High"
`time_slot_label`	meal_type + " " + hour_label	"Breakfast 08:00"
`priority_rank`	1 if High, 2 if Medium, 3 if Low	1

Troubleshooting & FAQs

Q: I get "File not found" when running module6

A: Ensure modules 2, 3, and 5 were executed first in order.

python module2_preprocessing.py  # Creates cleaned data
python module3_analytics.py      # Creates peak_hours, trends
python module5_recommendations.py # Creates recommendations
python module6_dashboard_prep.py  # Creates dashboard CSVs

Q: Module 6 runs but outputs are empty or missing

A: Check that all 10 CSV files exist in outputs/dashboard/:

dir outputs\dashboard\  # Windows
ls outputs/dashboard/   # Mac/Linux

Should show 10 files. If <10, check error messages in module output.

Q: Power BI relationships don't show up

A: Verify:

All 10 CSVs imported as tables (not queries)
Date columns are Date type, not Text
Hour columns are Whole Number, not Text
Column names match exactly

Q: DAX formula shows "Table of ..." or #ERROR

A:

Copy formula exactly (watch indentation & brackets)
Verify table/column names exist
Use SINGLE quotes for column names with spaces
Test with simple measures first

Q: Slicers don't filter charts

A:

Verify relationships exist in Model view
Check cross-filter direction is "Both"
Enable "Sync Slicers" across pages
Test with simplest slicer first

Q: Dashboard is slow/unresponsive

A:

Filter to smaller date range
Aggregate data weekly instead of daily
Use DirectQuery for large datasets
Reduce number of simultaneous visuals

Q: Anomaly page shows no data

A:

Verify is_anomaly field exists in electricity_clean.csv
Check that some rows have is_anomaly = 1
Apply filter: is_anomaly = 1 (not = 0)
Verify date range includes anomaly records

Next Steps

✅ Run Module 6 → Generate dashboard CSVs
✅ Choose BI Tool → Power BI (recommended) or Tableau
✅ Import Data → Import all 10 CSVs from outputs/dashboard/
✅ Create Relationships → Follow relationship diagram
✅ Add Measures → Paste DAX formulas (Power BI) or calculated fields (Tableau)
✅ Build Pages → Follow page-by-page guide in DASHBOARD_GUIDE.md
✅ Apply Theme → Use color palette & formatting specs
✅ Enable Slicers → Cross-page filtering
✅ Publish → Deploy to Power BI Service / Tableau Server
✅ Share → Distribute to stakeholders

Documentation Files

File	Purpose	Audience
README.md	System overview & quick start	Everyone
DASHBOARD_GUIDE.md	Complete BI implementation guide	BI Tool Builders
QUICK_REFERENCE.md	Printable color/formula reference	Developers
module6_dashboard_prep.py	Data preparation script	Data Engineers

Support Resources

Documentation:

DASHBOARD_GUIDE.md — Full step-by-step setup
QUICK_REFERENCE.md — Printable reference card
Code comments in all module files

Troubleshooting:

Check error messages in terminal output
Verify file paths and data types
Test formulas individually
Check DASHBOARD_GUIDE.md FAQ section

Performance Optimization:

Use date range filters to reduce data volume
Pre-aggregate data for large datasets
Enable query folding in Power BI
Use DirectQuery for real-time data

License & Attribution

This Smart Campus Intelligence System is developed for educational and enterprise use.

Key Technologies:

Python 3.8+ (pandas, numpy, scikit-learn)
Power BI Desktop / Tableau Desktop
SQL-based data warehousing (optional)

Data Privacy:

All student/personal data should be anonymized
Comply with GDPR, FERPA, and local regulations
Secure dashboard access with authentication

Version History

Version	Date	Changes
1.0	Apr 2026	Initial release with 6 modules, 10 dashboard tables, 6 BI pages

Contact & Feedback

For questions, bugs, or feature requests:

Check DASHBOARD_GUIDE.md Troubleshooting section
Review QUICK_REFERENCE.md for common issues
Inspect error messages in terminal output
Validate data in outputs/dashboard/ CSVs

Ready to build your dashboard? Start with:

python module6_dashboard_prep.py

Then open DASHBOARD_GUIDE.md for complete Power BI/Tableau setup instructions!

Smart Campus Intelligence System v1.0
April 2026
✅ Production Ready

Name		Name	Last commit message	Last commit date
Latest commit History 4 Commits
__pycache__		__pycache__
data		data
models		models
outputs		outputs
static		static
templates		templates
DASHBOARD_GUIDE.md		DASHBOARD_GUIDE.md
IMPLEMENTATION_SUMMARY.md		IMPLEMENTATION_SUMMARY.md
QUICK_REFERENCE.md		QUICK_REFERENCE.md
README.md		README.md
app.py		app.py
module1_data_generation.py		module1_data_generation.py
module2_preprocessing.py		module2_preprocessing.py
module3_analytics.py		module3_analytics.py
module4_ml_models.py		module4_ml_models.py
module5_recommendations.py		module5_recommendations.py
module6_dashboard_prep.py		module6_dashboard_prep.py
requirements.txt		requirements.txt
validate_project.py		validate_project.py

Folders and files

Latest commit

History

Repository files navigation

Smart Campus Intelligence System — Complete README

System Overview

Key Features

System Architecture

Module Hierarchy

Quick Start (5 Minutes)

Step 1: Run Data Pipeline

Step 2: Launch the Local Website

Step 3: Explore and Share

File Structure

Module Details

Module 2: Data Preprocessing & Cleaning

Module 3: Analytics & Aggregation

Module 4: ML Models (Optional)

Module 5: Recommendations

Module 6: Dashboard Preparation ⭐ THIS IS THE KEY

Dashboard Overview

6 Interactive Pages

📊 PAGE 1: Overview

🏗️ PAGE 2: Resource Utilization

⏰ PAGE 3: Peak Hours Analysis

📈 PAGE 4: Predictive Insights

🚨 PAGE 5: Anomaly Detection

💡 PAGE 6: Recommendations

Power BI Implementation Quick Reference

Create Relationships (Model View)

Paste Essential Measures (New Measure)

Global Slicers (Enable Cross-Page)

Tableau Implementation Quick Reference

Create Relationships (Data Source Tab)

Create Calculated Fields

Color Palette (Use Consistently)

Status Colors

Building Colors (Fixed)

Priority Colors (Recommendations Table)

Data Dictionary

Common Fields Across Tables

Calculated Columns (Added by Module 6)

Troubleshooting & FAQs

Q: I get "File not found" when running module6

Q: Module 6 runs but outputs are empty or missing

Q: Power BI relationships don't show up

Q: DAX formula shows "Table of ..." or #ERROR

Q: Slicers don't filter charts

Q: Dashboard is slow/unresponsive

Q: Anomaly page shows no data

Next Steps

Documentation Files

Support Resources

License & Attribution

Version History

Contact & Feedback

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages