Software Life Cycles & Process Models: Models, Phases & Workflows

TAKEAWAYS:

• The Software Development Life Cycle (SDLC) is a structured approach to planning, designing, building, testing, and deploying software, with 6 core phases: requirements, design, implementation, verification, maintenance, and retirement.
• Process models (waterfall, iterative, agile, spiral) define how work flows through SDLC phases—each has trade-offs between flexibility, cost, and risk.
• Waterfall is rigid and sequential; agile is iterative and adaptive. Choose based on project uncertainty, stakeholder needs, and budget constraints.
• Real-world tie: eSewa’s payment system uses an iterative model to handle frequent regulatory changes, while Daraz’s order queue relies on agile sprints for rapid feature updates.
• Critical failure point: Poor model selection (e.g., using waterfall for a research-heavy project) leads to scope creep, budget overruns, or abandoned projects (e.g., Nepal’s failed e-Governance portal in 2010).
• Exam focus: Compare models in tables, trace a project through phases, and justify model choices for given scenarios (e.g., "Why would Ncell use spiral over agile?").

1. The Software Development Life Cycle (SDLC): Phases and Workflow

The SDLC is a structured framework for developing software, ensuring systematic progress from idea to deployment. It consists of 6 phases, each with clear deliverables and review points. Visualizing SDLC as a cycle (not a linear process) helps understand its iterative nature.

stateDiagram-v2
    [*] --> Requirements:Gathering
    Requirements:Gathering --> System:Design
    System:Design --> Implementation:Development
    Implementation:Development --> Verification:Testing
    Verification:Testing --> Maintenance:Support
    Maintenance:Support --> [*]
    Maintenance:Support --> Retirement:Decommission

Key Phases Explained

Why SDLC Matters

• Reduces risks: Early phase reviews catch issues before coding (e.g., Ncell’s app failed its first launch due to unclear requirements).
• Controls costs: Prevents scope creep (e.g., Daraz’s initial MVP cost $500K; adding features later would have doubled it).
• Ensures quality: Structured testing (e.g., NEPSE’s stress tests) prevents crashes during high-volume trading.

A labeled flowchart of SDLC phases with arrows showing iteration loops. (Image: Cliffydcw, CC BY-SA 3.0, via Wikimedia Commons)

2. Process Models: How Work Flows Through SDLC

Process models define how the SDLC phases are executed. The choice of model impacts flexibility, cost, and risk. Below are the 4 primary models, compared in a table.

A. Waterfall Model: Sequential and Rigid

• Workflow: Linear, one-phase-to-the-next (no iteration).
• Best for: Projects with clear, stable requirements (e.g., government portals like e-Governance).
• Pros:
  • Simple to manage.
  • Clear milestones and documentation.
• Cons:
  • No going back (e.g., if requirements change in Phase 3, you must restart).
  • High risk of failure if requirements are unclear.

flowchart LR
    A["Requirements"] --> B["Design"]
    B --> C["Implementation"]
    C --> D["Testing"]
    D --> E["Maintenance"]

B. Iterative Model: Repeat Phases in Cycles

• Workflow: Repeat SDLC phases for partial deliverables (e.g., build a prototype, then refine).
• Best for: Projects needing early feedback (e.g., Pathao’s ride-hailing app).
• Pros:
  • Early testing reduces risks.
  • Adaptable to changes.
• Cons:
  • Higher initial cost (repeating phases).
  • Requires stakeholder availability.

flowchart TD
    A["Requirements"] --> B["Design"]
    B --> C["Implementation"]
    C --> D["Testing"]
    D -->|"Feedback"| A

C. Agile Model: Incremental and Collaborative

• Workflow: Work in sprints (2–4 weeks), with continuous stakeholder input.
• Best for: Dynamic projects (e.g., Daraz’s daily feature updates).
• Pros:
  • Fast delivery of working software.
  • High customer satisfaction.
• Cons:
  • Requires disciplined teams.
  • Documentation can lag.

flowchart LR
    A["Sprint 1"] --> B["Plan"]
    B --> C["Develop"]
    C --> D["Test"]
    D --> E["Review"]
    E -->|"Feedback"| A

D. Spiral Model: Risk-Driven Iterations

• Workflow: Combine iterative development with risk analysis in each cycle.
• Best for: High-risk, complex projects (e.g., Ncell’s 5G network rollout).
• Pros:
  • Identifies risks early.
  • Balances flexibility and control.
• Cons:
  • Expensive (requires risk analysis experts).
  • Overkill for small projects.

flowchart TD
    A["Plan"] --> B["Risk Analysis"]
    B --> C["Engineer"]
    C --> D["Evaluate"]
    D -->|"Feedback"| A

Comparison Table: Process Models

3. Choosing the Right Model: A Worked Example

Scenario: Nepal’s National Traffic Management Center (NTMC) wants to develop a real-time traffic monitoring system for Kathmandu.

Step 1: Analyze Project Characteristics

• Requirements: Unclear (sensors, AI, user apps).
• Stakeholders: NTMC, government, citizens.
• Budget: Limited ($2M).
• Risk: High (traffic data is volatile; regulations may change).

Step 2: Evaluate Models

Step 3: Recommendation

• Use Spiral Model:
  1. Cycle 1: Define core sensors + basic dashboard (risk: sensor accuracy).
  2. Cycle 2: Add AI for predictions (risk: false positives in alerts).
  3. Cycle 3: Integrate citizen feedback (risk: app usability).
• Why? Balances NTMC’s need for structure with the project’s high uncertainty.

4. Real-World Applications: Where These Models Are Used

A. eSewa’s Payment System: Iterative + Agile Hybrid

• Model Used: Iterative for core payment flow + agile for UI updates.
• Why?
  • Iterative: Built a prototype (2016) to test bank integrations before full rollout.
  • Agile: Daily sprints to fix bugs (e.g., failed transactions during Dashain).
• Outcome: 5M+ users; handles 10K transactions/minute.

B. Daraz’s Order Queue: Agile Sprints

• Model Used: Pure agile (2-week sprints).
• Why?
  • Dynamic: Competitors (Amazon, Flipkart) force rapid feature additions (e.g., "Cash on Delivery in 1 hour").
  • Customer feedback: Daily stand-ups with logistics teams to fix delays.
• Outcome: 95% order accuracy; 24/7 support.

C. Ncell’s 5G Rollout: Spiral Model

• Model Used: Spiral (4 cycles over 3 years).
• Why?
  • High risk: 5G requires new infrastructure (towers, spectrum licensing).
  • Risk analysis: Cycle 1 tested tower placement in Pokhara; Cycle 2 optimized for Kathmandu’s dense traffic.
• Outcome: 5G launched in 2023 with 90% coverage.

5. Common Pitfalls and How to Avoid Them

6. Exam Tip: How to Score Full Marks

1. Define clearly: Start with precise definitions (e.g., "The waterfall model is a sequential SDLC process where each phase must be completed before the next begins.").
2. Use tables for comparisons: Examiners love structured comparisons (e.g., waterfall vs. agile).
3. Tie to real-world examples: Always link models to Nepalese companies (e.g., "Like Daraz, projects with dynamic requirements should use agile.").
4. Trace a project through phases: For case studies, show how a model would handle each SDLC phase (e.g., "In the spiral model, Ncell’s 5G project would analyze risks in Cycle 1 before implementation.").
5. Highlight trade-offs: Explain why a model is chosen (e.g., "Waterfall is cheaper for stable projects but risky for eSewa’s evolving payment rules.").
6. Avoid vague answers: ❌ "Agile is flexible." ✅ "Agile’s 2-week sprints allow Daraz to update its order queue daily based on customer complaints."

Pro Tip: Memorize one Nepalese example per model (e.g., eSewa = iterative, Ncell 5G = spiral). Examiners love local context!