PROJECTS 

KPM V1 — Basic Motion (Failure-Driven Start)

• Built: Wheeled robot using Arduino Uno + L298N motor driver
  
  

• Problem: Robot repeatedly crashed into objects
  
  

• Lesson: Movement without sensing is useless → autonomy requires perception
  
  

• Programmed basic motion: forward, backward, left, and right

• Established initial wiring and control logic

 KPM V2 — Speed Experiment (Intentional Failure) Performance Upgrade

• Attempt: Added rear propulsion fan to increase speed
  
  

• Problem: High power drain, negligible speed gain
  
  

• Decision: Removed fan; Added a high-RPM rear propulsion fan to increase thrust

• Improved overall movement performance

• Lesson: Engineering values efficiency over brute-force solutions

KPM V3A — Rear Obstacle Detection Prototype

Upgrade: Added two rear-mounted infrared (IR) obstacle sensors to begin experimenting with obstacle detection.

Problem: The robot could only detect objects behind it and had no awareness of obstacles in front while moving.

Lesson Learned: Autonomous robots need sensors in the direction of travel. Rear sensing alone cannot provide reliable navigation.

Outcome: Successfully integrated basic obstacle detection and established the foundation for future sensor expansion.

KPM V3B — Front & Rear Detection System

Upgrade: Expanded the sensing system by adding front-mounted IR sensors while retaining the rear sensors.

Problem: Although front and rear detection improved obstacle awareness, blind spots still remained due to the limited sensor angles.

Lesson Learned: Increasing the number of sensors improves environmental awareness, but strategic sensor placement is equally important.

Outcome: Achieved front and rear obstacle detection, leading to experiments with alternative sensor orientations.

KPM V3C — Multi-Directional Sensor 

Upgrade: Added an additional IR sensor facing straight upward while keeping the front and rear sensors, creating a three-direction sensing system.

Problem: Standard front and rear sensors still left areas of limited visibility and encouraged testing of alternative sensing angles.

Lesson Learned: Effective autonomous navigation depends on both the number of sensors and their orientation. Different sensor angles provide different perspectives of the environment and improve overall situational awareness.

Outcome: Successfully tested a multi-directional sensing layout, providing valuable insights that influenced later KPM robot designs and future autonomous navigation systems.

KPM V4A — 360° Scanning Prototype

Upgrade: Introduced a servo-mounted ultrasonic sensor with a full 360° scanning system to give the robot complete environmental awareness.

Problem: While the robot could scan in every direction, the continuous 360° rotation caused slow response times, making obstacle avoidance less effective during movement.

Fix: Evaluated the system and determined that full-circle scanning was unnecessary for forward navigation.

Lesson Learned: More sensing coverage does not always result in better performance. Engineering requires balancing coverage, reaction speed, and efficiency.

Outcome: The prototype demonstrated the advantages and limitations of full environmental scanning, leading to a more optimized navigation system.

KPM V4B — Optimized Hybrid Navigation System

Upgrade: Reduced the ultrasonic sensor's scanning range from 360° to 180° for faster obstacle detection. Added two front IR sensors and two rear IR sensors to create a multi-layer obstacle detection system, similar to the proximity sensors used in modern vehicles.

Problem: The 360° system reacted too slowly for real-time navigation and still required additional close-range obstacle detection.

Fix: Combined a 180° ultrasonic scanning system with dedicated front and rear IR safety sensors. The ultrasonic sensor handled long-range scanning while the IR sensors provided immediate obstacle detection.

Lesson Learned: Combining multiple sensor technologies creates a faster, more reliable autonomous navigation system than relying on a single sensor. Optimized sensor placement and redundancy significantly improve real-world performance.

Outcome: Developed a hybrid navigation system featuring a 180° ultrasonic scanner with front and rear IR safety sensors, resulting in quicker reactions, improved reliability, and smoother autonomous navigation.

KPM V5 — Reliability & Real-World Improvements

Power Redesign: Rewired the robot to use two separate rechargeable 9V battery systems:

• Battery 1: Powered the Arduino, sensors, and LED lighting.

• Battery 2: Dedicated exclusively to the drive motors.

Result: Isolating the power systems eliminated voltage drops caused by the motors, reduced electrical interference with the sensors, and provided more consistent motor performance and speed.

Physical Redesign: Lowered the front chassis to improve obstacle detection and added LED lighting for better operation in low-light environments.

Lesson Learned: Separating high-current motor loads from sensitive electronics significantly improves reliability, sensor accuracy, and overall robot performance. Proper power management is just as important as programming in autonomous robotics.

Outcome: Developed a stable and reliable autonomous robot featuring isolated power systems, improved sensor performance, smoother motor operation, enhanced low-light visibility, and the use of rechargeable batteries for extended operation.

KPM V6 — Advanced Modular Recon Robot ( Still building Current) 

• Current Work: Design-focused iteration
  
  

• Goals:
  
  

  • 4-wheel drive platform
    
    

  • Quick-swap base (wheels ↔ caterpillar tracks)
    
    

  • Modular components inspired by F1-style fast replacement
    
    

  • Camera module planned
    
    

  • Upgraded to a 3D-printed chassis for strength and precision

  • Integrated a rotating camera for environmental monitoring

  • Added dual control systems:

    1. Bluetooth smartphone control

    2. Long-range infrared control

• Focus: Maintainability, usability, and rapid repair

Multi-Sensor Safety Glove (Human-Centered Design Project)

Initial Concept — Smart Jacket (Failed Prototype)

• Designed a full wearable jacket with embedded sensors

• Identified major issues:

  1. Fragile wiring

  2. Difficult to wear and use quickly

• Not practical for real-world situations
  
  

Final Design — Safety Glove

Features:

• IR sensors on fingertips for directional sensing

• Ultrasonic sensor on back of hand for distance detection

• Flame sensor on pinky for heat detection

Key Improvements:

• More natural interaction using hand movement

• Stronger and more reliable wiring

• Designed for real-world use instead of just a prototype

• Outcome: Fast to deploy, durable, intuitive UX

• Lesson: Technology must adapt to users, not the other way around