IoT Basics: Definitions, M2M vs IoT, Architecture & Real-World Impact

1. Definitions: What Is IoT and M2M?

IoT (Internet of Things) refers to a network of physical objects (things) embedded with sensors, software, and connectivity to collect and exchange data. M2M (Machine-to-Machine) is a subset of IoT where machines communicate without human interaction.

Key Differences

2. How IoT Works: The 5-Layer Architecture

IoT systems are built on 5 layers, each with distinct roles. Below is the Mermaid diagram of the architecture, followed by a breakdown:

Layer Breakdown

1. Perception Layer

   • Devices: Sensors (temperature, motion), actuators (motors, LEDs).
   • Example: A smart traffic light (sensor detects cars → actuator adjusts timing).
   • IMAGE: traffic light controller circuit | Real IoT sensor-actuator setup for Kathmandu’s smart signals.

2. Network Layer

   • Protocols: Wi-Fi, LoRaWAN, NB-IoT, or IPv6 (critical for 65,000+ devices).
   • Example: NTC’s smart meters use LoRaWAN to send usage data to the cloud without Wi-Fi.

3. Processing Layer

   • Edge vs. Cloud:
     • Edge: Local processing (e.g., Google Nest adjusting thermostats before cloud sync).
     • Cloud: Centralized analytics (e.g., WhatsApp’s server farms processing messages).

4. Application Layer

   • Interfaces: Mobile apps (eSewa), dashboards (Daraz logistics), or APIs.
   • Example: Khalti’s fraud detection uses IoT data (transaction speed, location) to flag anomalies.

5. Business Layer

   • Outcomes: Revenue models (subscription, ads), ROI analysis.
   • Example: NEPSE’s stock market IoT sensors predict trading patterns via predictive analytics.

3. M2M Communication: Steps and Example

M2M follows a 4-step cycle:

1. Data Collection: Machine (e.g., ATM) reads card → sends transaction data.
2. Processing: Bank server validates transaction (no human).
3. Action: ATM dispenses cash or declines.
4. Feedback: Bank sends confirmation to merchant’s system.

Worked Example: Pathao’s Ride Optimization

• Step 1: Driver’s phone (sensor) detects idle time.
• Step 2: M2M message sent to Pathao’s server via NB-IoT.
• Step 3: Server reroutes driver to nearest passenger (no human dispatch).
• Step 4: Passenger’s app updates ETA in real-time.

sequenceDiagram
    Driver->>+Server: [M2M] "Idle detected (GPS)"
    Server->>-Driver: [M2M] "Route to Passenger X"
    Passenger->>Server: [App] "Request ride"
    Server-->>Passenger: [App] "Driver Y assigned"

4. Arduino vs. Raspberry Pi for IoT

When to Use Which?

• Arduino: Low-power, single-task devices (e.g., NTC’s water meter sensors).
• Raspberry Pi: Multi-task systems (e.g., Daraz’s warehouse inventory trackers).

5. Real-World IoT in Nepal and Globally

Nepal

1. eSewa

   • Idea Used: M2M + Cloud Processing
   • How: When you pay a bill via eSewa, the app sends an M2M request to the utility’s server (e.g., NTC). The server validates your account → updates the bill status → sends a confirmation SMS (no human intervention).

2. NTC Smart Meters

   • Idea Used: IPv6 + LoRaWAN
   • How: Each meter has an IPv6 address (e.g., 2001:db8::1234). Data travels via LoRaWAN (long-range, low-power) to NTC’s cloud, where analytics predict peak usage hours.

3. Khalti Fraud Detection

   • Idea Used: Edge Analytics
   • How: If a transaction seems unusual (e.g., sudden large amount in Pokhara from Kathmandu), Khalti’s edge server flags it before cloud processing.

Global

1. Google Nest Thermostat

   • Idea Used: IoT + Machine Learning
   • How: Learns your schedule (via phone GPS) → adjusts temperature before you arrive home (edge computing).

2. WhatsApp’s Message Routing

   • Idea Used: M2M + Load Balancing
   • How: When you send a message, WhatsApp’s servers use M2M to route it to the recipient’s device via the fastest path (no human relay).

3. Tesla’s Over-the-Air Updates

   • Idea Used: Firmware Updates (IoT Layer 5)
   • How: Tesla cars receive automatic software patches via cellular M2M, fixing bugs without dealership visits.

6. Why IPv6 Matters for IoT

• Problem: IPv4 only supports ~4.3 billion addresses (Nepal has ~30M people + devices).
• Solution: IPv6 offers 340 undecillion addresses (enough for 65,000 devices per square meter on Earth).
• Example: NTC’s smart grid uses IPv6 to assign unique IDs to every meter, enabling real-time monitoring.

IPv6 Packet Format (vs. IPv4):

Exam Tip

1. Definitions First: Always start with IoT = sensors + connectivity + data → action. M2M is subset of IoT (no humans).
2. Layer Questions: For the 5-layer architecture, describe one function per layer + one Nepalese/global example.
   • Example Answer:

     "The Network Layer in NTC’s smart meters uses LoRaWAN to transmit data over long distances with low power, ensuring rural areas stay connected."

3. Comparisons: For Arduino vs. Pi, use the table above and add a use case (e.g., "Arduino is ideal for Pathao’s bike GPS trackers due to its low cost").
4. Real-World Links: Always tie answers to Nepal (eSewa, NTC, Khalti) or global tech (Google Nest, WhatsApp). Examiners love this!
5. Diagrams: Draw the 5-layer architecture or M2M sequence in exams—it’s worth 2+ marks and shows understanding.

Final Checklist Before Exam: ✅ Can you draw the 5-layer IoT model? ✅ Do you know 2 Nepalese IoT examples (eSewa, NTC, Khalti)? ✅ Can you compare Arduino vs. Raspberry Pi in 3 points? ✅ Do you understand IPv6’s role in scaling IoT?