SystemIQ — Intelligent PC Hardware Troubleshooter

SystemIQ is an interactive, explainable expert system for diagnosing PC hardware and boot issues using a knowledge base of facts, rules, and confidence factors. The system uses intelligent question adaptation to minimize unnecessary questions and provide accurate diagnostics efficiently.

Features

Smart Question Adaptation

SystemIQ features two diagnostic modes:

• Smart Mode (Default): Intelligently skips irrelevant questions based on previous answers

  • If power LED is off, skips questions about BIOS, OS, and other components
  • If BIOS is inaccessible, skips questions about BIOS-visible hardware status
  • If no display output, skips questions about visual artifacts
  • If OS doesn't load, skips runtime-specific questions
  • And more conditional logic to streamline diagnostics

• All Mode: Asks all questions for comprehensive data collection

  • Useful for thorough documentation or edge cases
  • Provides complete dataset for knowledge base refinement

Confidence-Based Ranking

• Diagnoses are ranked by confidence factor (CF)
• Each diagnosis includes the triggering rule ID and explanation
• Multiple diagnoses may be returned when symptoms overlap

JSON-Driven Knowledge Base

• Easily extensible without modifying Python code
• Facts define diagnostic questions and valid answers
• Rules define condition-diagnosis mappings with confidence factors
• Clean separation between logic and domain knowledge

Installation

For CLI Usage (No Dependencies)

The command-line interface uses only Python standard library:

# Clone or download the repository
cd SystemIQ

# Run directly
python systemiq_cli.py

For Web Interface

The web interface requires Flask:

# Clone or download the repository
cd SystemIQ

# Install dependencies
pip install -r requirements.txt

# Run the web server
python app.py

Then open your browser to http://localhost:5000

Usage

Web Interface (Recommended)

The modern web interface provides a sleek, interactive experience:

1. Start the server:

   python app.py

2. Open your browser to http://localhost:5000

3. Choose your mode:

   • Smart Mode: Adaptive questions that skip irrelevant diagnostics
   • Complete Mode: All questions for comprehensive analysis

4. Answer questions by clicking on the answer buttons

5. View results with confidence-ranked diagnoses and explanations

Features:

• Modern, responsive UI with dark theme and gradient effects
• Real-time progress tracking
• Confidence factor visualization
• Detailed explanations for each diagnosis
• Easy reset for multiple diagnostics
• Responsive design for mobile and desktop

Command-Line Interface

Run SystemIQ in smart mode (default):

python systemiq_cli.py

Command-Line Options

# Smart mode (adaptive questions)
python systemiq_cli.py smart
python systemiq_cli.py --smart
python systemiq_cli.py -s

# All mode (ask all questions)
python systemiq_cli.py all
python systemiq_cli.py --all
python systemiq_cli.py -a

# Help
python systemiq_cli.py --help
python systemiq_cli.py -h

Example Session (Smart Mode)

$ python systemiq_cli.py

Welcome to SystemIQ — Intelligent PC Hardware Troubleshooter!
Running in SMART mode: Questions adapt based on your answers.
Answer the questions to receive ranked diagnoses with explanations.

Can you enter BIOS/UEFI setup? (yes/no): no
Is the power LED on? (yes/no): yes
Do you hear BIOS beep codes? (none/short/long/repeated): short
Are system fans spinning? (normal/slow/stopped): normal
Does the PC overheat or shut down under load? (yes/no): yes
Do you see visual artifacts/flickering on display? (yes/no): yes
Are fan noises unusually loud at idle? (yes/no): yes

Evaluating knowledge base...

Diagnosis ranking:
- gpu_overheating_or_vram_fault: CF=0.89 (via R10)
- motherboard_failure: CF=0.85 (via R4)
- dust_buildup_or_bearing_wear: CF=0.72 (via R18)

Explanations:
- R10: Visual artifacts combined with heat often indicate GPU thermal or VRAM issues.
- R4: Short POST beeps with no BIOS access suggest motherboard fault.
- R18: Normal RPM but loud fan noise usually indicates dust buildup or worn bearings.

Notice how in the example above, smart mode automatically skipped questions about:

• BIOS-visible hardware status (RAM, HDD detection) since BIOS is inaccessible
• OS loading and network connectivity (since we can't even reach BIOS)
• Boot device detection (requires BIOS access)

Example Session (All Mode)

$ python systemiq_cli.py all

Welcome to SystemIQ — Intelligent PC Hardware Troubleshooter!
Running in ALL mode: All questions will be asked for comprehensive analysis.
Answer the questions to receive ranked diagnoses with explanations.

[All diagnostic questions will be asked regardless of previous answers]

Knowledge Base Structure

The knowledge base is located at knowledge/knowledge_base.json and consists of two main sections:

Facts Section

Defines diagnostic facts (questions) and their valid answers:

{
  "facts": {
    "power_led": {
      "question": "Is the power LED on? (yes/no): ",
      "answers": ["yes", "no"]
    },
    "bios_access": {
      "question": "Can you enter BIOS/UEFI setup? (yes/no): ",
      "answers": ["yes", "no"]
    }
    ...
  }
}

Rules Section

Defines diagnostic rules with conditions, diagnosis labels, confidence factors, and explanations:

{
  "rules": [
    {
      "id": "R1",
      "conditions": {"power_led": "no"},
      "diagnosis": "power_supply_failure",
      "cf": 0.96,
      "explanation": "No power LED strongly indicates PSU failure or no PSU output."
    }
    ...
  ]
}

Smart Mode Logic

The smart mode implements conditional question dependencies to avoid asking illogical questions:

Previous Answer	Questions Skipped	Reason
power_led: no	Most hardware questions	PC has no power - most diagnostics irrelevant
bios_access: no	ram_status, hdd_status	Cannot view hardware in BIOS
display_output: no	artifacts	Cannot see visual artifacts without display
os_loads: no	blue_screen (runtime)	OS never starts, so no runtime crashes
fan_speed: stopped	high_noise	Fans not spinning, so no noise
os_loads: no	network_link	Network requires OS to be running
bios_access: no	usb_devices_fail (if OS also fails)	Cannot test USB without BIOS or OS

This logic can reduce question count by 30-50% while maintaining diagnostic accuracy.

Extending the System

Adding New Facts

1. Open knowledge/knowledge_base.json
2. Add a new fact definition in the facts section:

"new_fact_name": {
  "question": "Your diagnostic question? (answer1/answer2): ",
  "answers": ["answer1", "answer2", "answer3"]
}

Adding New Rules

Add a new rule in the rules section:

{
  "id": "R21",
  "conditions": {
    "fact1": "value1",
    "fact2": "value2"
  },
  "diagnosis": "diagnosis_label",
  "cf": 0.85,
  "explanation": "Why these conditions suggest this diagnosis."
}

Tips for effective rules:

• Use confidence factors (cf) between 0.0 and 1.0
• Higher CF = stronger certainty of diagnosis
• Combine multiple symptoms for more specific diagnoses
• Write clear, actionable explanations

Adjusting Smart Mode Logic

To modify question dependencies, edit the should_ask_question() function in systemiq_cli.py:

def should_ask_question(facts: dict[str, str | None], fact: str, smart_mode: bool) -> bool:
    """Determine if a question should be asked based on previous answers."""
    if not smart_mode:
        return True
    
    # Add your conditional logic here
    if facts.get("some_fact") == "some_value":
        if fact in ["skip_this", "and_this"]:
            return False
    
    return True

Technical Architecture

Inference Engine

• Forward chaining inference: Questions are selected based on which facts could still satisfy unmatched rules
• Prioritized question selection: Facts that appear in more potential rules are asked first
• Dynamic question filtering: Smart mode skips questions that are logically impossible or irrelevant

Confidence Factors

• Each rule has a CF between 0.0 and 1.0 indicating diagnostic certainty
• Multiple rules can fire for the same diagnosis
• Results are ranked by confidence for clear prioritization

Knowledge Representation

• Facts: Observable symptoms and hardware states
• Rules: IF-THEN mappings from symptom combinations to diagnoses
• Explanations: Human-readable justifications for each diagnosis

Diagnostic Categories

SystemIQ can diagnose issues across multiple hardware categories:

• Power System: PSU failures, power delivery issues
• Cooling: Fan failures, dust buildup, thermal issues
• Memory: RAM failures, seating issues, POST errors
• Storage: Drive failures, connection issues, bootloader problems
• Display: GPU failures, VRAM issues, display path problems
• Motherboard: Chipset issues, POST failures, USB controller problems
• Operating System: Boot failures, driver conflicts, corruption
• Network: Adapter failures, driver issues, connectivity problems

Use Cases

For Technicians

• Structured diagnostic workflow
• Documentation of troubleshooting steps
• Training tool for new technicians
• Consistent diagnostic approach

For End Users

• Self-diagnosis before seeking professional help
• Understanding potential hardware issues
• Prioritizing repair actions by confidence

For Researchers

• Knowledge base for hardware failure patterns
• Testing ground for expert system techniques
• Dataset for diagnostic AI training

Contributing

To improve SystemIQ:

1. Test diagnostics against real hardware issues
2. Refine confidence factors based on real-world accuracy
3. Add new rules for additional failure modes
4. Enhance conditional logic for smarter question flow
5. Expand explanations for better user understanding
6. Document edge cases where the system needs improvement

Known Limitations

• Does not diagnose software-specific issues beyond boot failures
• Cannot detect intermittent hardware failures
• Requires user to accurately assess hardware symptoms
• Limited to common PC hardware configurations
• May produce multiple diagnoses with similar confidence

Future Improvements

• Web-based interface for remote diagnostics
• Diagnostic history and session logging
• Machine learning to refine confidence factors
• Integration with hardware monitoring tools
• Multi-language support
• Cloud-based knowledge base sharing

License

This project is open source and available for educational and diagnostic purposes.

Acknowledgments

Built using expert system principles and forward-chaining inference techniques. Knowledge base compiled from common PC hardware troubleshooting scenarios.

---

SystemIQ - Making PC diagnostics smarter, one question at a time.