Power Sources & Sensors: Types, Working, and Applications

TAKEAWAYS:

• Robots need energy sources (primary/secondary) and sensors (contact/non-contact) to operate autonomously, with trade-offs between cost, efficiency, and environmental impact.
• Batteries (Li-ion, NiMH) dominate mobile robots, while fuel cells and solar panels enable long-duration tasks like agricultural drones in Nepal’s hills.
• Proximity/ultrasonic sensors detect obstacles in tight spaces (e.g., Pathao’s delivery robots), while LiDAR/IMU enable precise navigation (e.g., NTC’s automated traffic monitoring).
• Sensor fusion combines data (e.g., GPS + IMU) to improve accuracy, critical for applications like NEPSE’s automated stock market surveillance drones.
• Calibration and noise reduction are essential for reliable sensor readings, often tested in exams via signal processing questions.
• Ethical/safety considerations (e.g., sensor privacy in surveillance) are increasingly relevant in automated systems.

---

1. Power Sources for Robots and Automated Systems

Robots require reliable, efficient, and often portable power sources. These can be categorized into primary (non-rechargeable) and secondary (rechargeable) sources, with trade-offs in energy density, lifespan, and cost.

1.1 Primary Power Sources

These are single-use or require external recharging but offer high energy density.

1.1.1 Batteries (Primary)

• Types:
  • Alkaline: Cheap, widely used in toys/remote controls (e.g., Daraz’s warehouse robots).
  • Lithium (Li-SO₂): High energy density, used in drones (e.g., Pathao’s delivery drones).
  • Zinc-Air: Long shelf life, used in hearing aids.
• Pros: Lightweight, no maintenance.
• Cons: Disposable, environmental hazards (e.g., mercury in older types).
• Worked Example: A Li-SO₂ battery in a drone has a capacity of 5000 mAh and supplies 3.6V. If the drone’s motor draws 2A, how long will it run? Solution: $$\text{Time}=\frac{\text{Capacity (Ah)}}{\text{Current (A)}}=\frac{5000\text{ mAh}}{2000\text{ mA}}=2.5\text{ hours}$$ Real-world tie: Pathao’s drones use Li-ion variants for ~30-minute flights per charge.

1.1.2 Fuel Cells

• How it works: Converts chemical energy (e.g., hydrogen + oxygen) into electricity via an electrochemical reaction.
• Types:
  • Proton Exchange Membrane (PEM): Used in forklifts, military robots.
  • Solid Oxide Fuel Cells (SOFC): High efficiency, used in stationary power.
• Pros: High energy density, refuelable, zero emissions (if hydrogen is green).
• Cons: Expensive, requires hydrogen infrastructure.
• Nepal Example: NTC is piloting hydrogen fuel cells for electric buses in Kathmandu to reduce pollution.

1.2 Secondary Power Sources (Rechargeable)

These are reusable and dominate modern robotics.

1.2.1 Batteries (Secondary)

• Lead-Acid: Cheap, used in forklifts (e.g., Daraz warehouses).
• Nickel-Metal Hydride (NiMH): Higher energy than lead-acid, used in older robots.
• Lithium-Ion (Li-ion): Most common in drones, robots (e.g., Boston Dynamics’ Spot).
  • Types: LiCoO₂, LiFePO₄ (safer, used in Ncell’s solar-powered base stations).
• Pros: High energy density, lightweight, rechargeable.
• Cons: Degrades over time, fire risk (e.g., recall of Samsung Galaxy Note 7 batteries).
• Worked Example: A Li-ion battery in a warehouse robot has a 100Wh capacity and supplies 48V. If the robot’s motors draw 5A, how far can it travel if its efficiency is 80% and it needs 10W per meter? Solution:
  1. Total energy available = 100Wh = 360,000 J.
  2. Useful energy = 360,000 J × 0.8 = 288,000 J.
  3. Energy per meter = 10W × 1s = 10J (assuming 1s per meter).
  4. Distance = 288,000 J / 10 J/m = 28,800 meters (28.8 km). Real-world tie: Daraz’s warehouse robots use Li-ion for 8-hour shifts (~10 km/day).

1.2.2 Supercapacitors

• How it works: Stores energy electrostatically (double-layer capacitance).
• Pros: Fast charging, long lifespan (~1 million cycles), high power density.
• Cons: Low energy density (used for bursts, not sustained power).
• Example: Used in regenerative braking systems (e.g., NTC’s electric buses).

1.2.3 Solar Panels

• How it works: Photovoltaic cells convert sunlight into DC electricity.
• Pros: Renewable, low maintenance, ideal for outdoor robots (e.g., agricultural drones).
• Cons: Weather-dependent, low efficiency (~15-20%).
• Nepal Example: Nepal Electricity Authority (NEA) uses solar-powered drones to inspect power lines in remote areas like Mustang.

1.3 Comparison Table: Power Sources

---

2. Sensors in Automation and Robotics

Sensors provide robots with perception of their environment. They can be classified into:

1. Contact Sensors: Physical interaction (e.g., limit switches).
2. Non-Contact Sensors: Remote sensing (e.g., LiDAR, cameras).

2.1 Contact Sensors

Detect physical properties like pressure, force, or proximity via touch.

2.1.1 Limit Switches

• How it works: A mechanical switch that opens/closes when an object touches it.
• Types:
  • Normally Open (NO): Closed when activated.
  • Normally Closed (NC): Open when activated.
• Example: Used in conveyor belts (e.g., Daraz’s packaging robots) to detect item presence.
• Worked Example: A limit switch in a robotic arm’s gripper is NO. If the gripper closes and the switch opens, what does this indicate? Solution: The gripper has not detected the object (switch should close when object is gripped). This could mean:
  • Object is missing.
  • Gripper is misaligned.
  • Switch is faulty.

2.1.2 Force/Torque Sensors

• How it works: Measures applied force or torque (e.g., strain gauges, piezoelectric sensors).
• Example: Used in surgical robots (e.g., Ncell’s telemedicine robots) to ensure precise tool handling.

2.2 Non-Contact Sensors

Detect properties without physical contact, enabling remote sensing.

2.2.1 Proximity Sensors

• How it works: Detects nearby objects using electromagnetic fields, light, or sound.
• Types:
  • Inductive: Metal detection (e.g., in assembly lines).
  • Capacitive: Non-metal detection (e.g., liquid level sensors).
  • Optical: Uses infrared or laser (e.g., Pathao’s delivery robots avoid obstacles).
• Example: NTC’s automated toll booths use inductive sensors to detect vehicles.

2.2.2 Ultrasonic Sensors

• How it works: Emits high-frequency sound waves (20kHz–200kHz) and measures echo time to detect distance.
• Formula: $$\text{Distance}=\frac{\text{Speed of Sound}\times \text{Time}}{2}$$ (Divide by 2 because the sound travels to the object and back.)
• Worked Example: An ultrasonic sensor emits a sound wave at 343 m/s (speed of sound in air). The echo returns after 0.05 ms. What is the distance to the object? Solution: $$\text{Distance}=\frac{343\text{ m/s}\times 0.00005\text{ s}}{2}=0.008575\text{ m}\approx 8.6\text{ cm}$$ Real-world tie: Pathao’s delivery robots use ultrasonic sensors to navigate narrow alleys in Kathmandu.

2.2.3 LiDAR (Light Detection and Ranging)

• How it works: Uses laser pulses to create 3D maps of the environment.
• Types:
  • Mechanical: Rotating laser (e.g., Velodyne LiDAR).
  • Solid-State: No moving parts (e.g., used in self-driving cars).
• Example: NTC’s traffic monitoring drones use LiDAR to count vehicles and detect accidents.
• Visual Output:

2.2.4 IMU (Inertial Measurement Unit)

• How it works: Combines accelerometers (measure linear acceleration) and gyroscopes (measure angular velocity) to track orientation.
• Example: Used in drones (e.g., Pathao’s delivery drones) for stabilization.
• Worked Example: An IMU in a drone measures:
  • Acceleration: 2 m/s² (forward).
  • Angular velocity: 0.5 rad/s (yaw). If the drone starts from rest, what is its velocity after 3 seconds? Solution: $$v=u+at=0+2\text{ m/s²}\times 3\text{ s}=6\text{ m/s}$$ Real-world tie: Ncell’s drone network uses IMUs to maintain stable connections during flights.

2.2.5 Cameras (Computer Vision)

• How it works: Captures visual data, processed via image processing (edge detection, object recognition).
• Types:
  • RGB Cameras: Color images (e.g., YouTube’s live streaming).
  • Depth Cameras: Measures distance (e.g., Microsoft Kinect).
• Example: NEPSE’s automated surveillance uses cameras to detect fraudulent trading activity.
• Visual Output:

2.2.6 Sensor Fusion

Combines data from multiple sensors (e.g., GPS + IMU + LiDAR) to improve accuracy.

• Example: Self-driving tuk-tuks (piloted by Nepalese startups) use sensor fusion to navigate Kathmandu’s chaotic traffic.
• Worked Example: A robot’s position is estimated by:
  • GPS: ±5 m error.
  • IMU: ±0.1 m/s² drift.
  • LiDAR: ±0.05 m error. Which sensor is most reliable for short-term navigation? Solution: LiDAR (highest precision for short distances). GPS is better for long-range but less accurate locally.

2.3 Sensor Calibration and Noise Reduction

Sensors must be calibrated (adjusted for accuracy) and filtered (to reduce noise).

2.3.1 Calibration

• Example: An ultrasonic sensor reads 10 cm when the actual distance is 12 cm. The offset error is 2 cm.
• Solution: Apply a correction factor: $$\text{Actual Distance}=\text{Measured Distance}+2\text{ cm}$$

2.3.2 Noise Reduction

• Methods:
  • Moving Average Filter: Smooths rapid fluctuations.
  • Kalman Filter: Predicts and corrects sensor errors (used in drones).
• Example: A temperature sensor in a robotic arm reads 25.3°C, 24.8°C, 25.5°C. The moving average (3-sample) is: $$\frac{25.3+24.8+25.5}{3}=25.2°C$$

---

## In the Real World

1. Pathao’s Delivery Robots

   • Power Source: Li-ion batteries (rechargeable, ~30-minute runtime).
   • Sensors:
     • Ultrasonic: Avoids pedestrians in narrow alleys.
     • IMU: Maintains balance on uneven pavements.
     • GPS: Navigates to delivery points via shortest path.

2. NTC’s Automated Traffic Monitoring

   • Power Source: Solar panels + supercapacitors (for peak power during blackouts).
   • Sensors:
     • LiDAR: Counts vehicles and detects accidents on busy routes (e.g., Ring Road).
     • Cameras: Computer vision for license plate recognition (used in e-tolling).

3. NEPSE’s Stock Market Surveillance

   • Power Source: Backup Li-ion batteries (for 24/7 operation).
   • Sensors:
     • High-speed cameras: Detect unusual trading patterns (e.g., spoofing).
     • IMU + GPS: Ensures drones remain stable during high winds in Kathmandu’s valley.

---

## Exam Tip

This unit is heavily tested on:

1. Calculations:
   • Battery runtime (Ah × V / W).
   • Sensor distance (ultrasonic time-of-flight).
   • Energy efficiency (Wh → J → distance).
2. Comparisons:
   • Power source trade-offs (e.g., "Why use Li-ion over lead-acid?").
   • Sensor suitability (e.g., "Which sensor for a dark warehouse?").
3. Applications:
   • Real-world examples (e.g., "How does Pathao use ultrasonic sensors?").
   • Ethical/safety concerns (e.g., "Privacy risks of LiDAR in cities").
4. Diagrams:
   • Draw a fuel cell reaction or sensor fusion pipeline.
   • Label a LiDAR point cloud or IMU axes.

Common Pitfalls:

• Forgetting to divide by 2 in ultrasonic distance calculations.
• Confusing energy density (Wh/kg) with power density (W/kg).
• Ignoring sensor noise in real-world scenarios (examiners love this!).

High-Score Strategy:

• Always include units in calculations (e.g., "5000 mAh" not just "5").
• Relate to Nepal (e.g., "NTC uses LiDAR for traffic...").
• Draw diagrams for sensor fusion or power source comparisons.