A custom web-based Preventive Maintenance System developed for PT Excelitas Batam to streamline and automate maintenance scheduling and tracking across multiple machines, ensuring minimal downtime and optimal operational efficiency.

Key Features:

• Machine Management: Each machine in the system is assigned a designated engineer or person in charge (PIC), with detailed scheduling and notification features to enhance accountability and clarity.
• Flexible Scheduling & Intervals: Maintenance schedules can be set at specific intervals for each machine, making it easy to track and execute preventive measures.
• Customizable Form Templates: Customers can create unique templates for each maintenance schedule, allowing flexibility and specificity in documenting tasks for different machines.
• Approval Workflow: Each maintenance schedule includes an assigned approval person to review and validate task completion, maintaining high operational quality and safety standards.
• Automated Notifications: Engineers receive alerts when a maintenance task is due, ensuring timely attention to each schedule.
• Spare Parts Management: Each machine has an associated list of spare parts. The system monitors stock levels and notifies the engineer when spare part stock falls below safe levels, ensuring continuous operation without unexpected downtimes.

Technologies Used:

The system is built using Golang for the backend to ensure efficient processing and performance, and React for the frontend to provide a responsive and interactive user interface. This combination of technologies results in a scalable, high-performance solution that meets the preventive maintenance needs of PT Excelitas Batam.

SIBI Hand Gesture Detection

This project, developed as part of an undergraduate thesis, involves creating a machine learning model capable of recognizing static hand gestures in the SIBI (Sistem Isyarat Bahasa Indonesia) alphabet, enhancing communication accessibility for the Indonesian Sign Language community. Built using TensorFlow and MediaPipe Hands as the base model, the system effectively identifies hand gestures with high accuracy and precision.

Model Performance:

• Accuracy: 91.12%
• Precision: 0.9188
• Recall: 0.9111
• F1-Score: 0.9100

These metrics demonstrate the model's reliable performance in recognizing static hand gestures, making it well-suited for practical applications in assistive technologies.

Deployment & Interaction:

The model is deployed as a web-based interactive application, allowing users to test and interact with the SIBI hand gesture recognition system in real-time. Python, Socket.IO, and Flask power the backend, ensuring responsive, low-latency communication between the model and the web interface.

Technologies Used:

The machine learning model leverages TensorFlow and MediaPipe Hands for gesture detection, while Python, Socket.IO, and Flask enable an interactive, web-based deployment.

This undergraduate thesis project provides an efficient, accurate tool for SIBI hand gesture recognition, offering a powerful aid for communication in the Indonesian sign language community.

Muhasabah Android App

This Android application helps users build positive habits and monitor behaviors over time by recording both good and bad habits. Developed using Java, the app provides a simple, intuitive platform for users to set personal goals, track activities, and visualize progress through interactive graphs.

Key Features:

• Custom Habit Tracking: Users can define their own good and bad habits, tailoring the app to fit their unique routines and self-improvement goals.
• Activity Logging: Each time a habit is performed, users can record it in the app, capturing the frequency of each habit and reinforcing accountability.
• Progress Graphs: Users can view charts displaying their habit trends over time, allowing them to identify patterns, track improvements, and stay motivated.

Technology Used:

This project was developed using Java for Android, providing a responsive and user-friendly experience for mobile users.

This Habit Tracker app is a practical tool for self-reflection and personal growth, empowering users to cultivate positive routines and manage their habits effectively.

Download the app on Google Play Store

VTOL Vision and Control System for Kontes Robot Terbang Indonesia 2019

As part of the GMFC team representing Gadjah Mada University in the 2019 Kontes Robot Terbang Indonesia (KRTI), I was responsible for developing the vision and control programming for a VTOL (Vertical Take-Off and Landing) robot. Competing in the VTOL category, our team earned a distinguished 2nd place at KRTI 2019.

Project Responsibilities:

• Vision Programming: Using OpenCV, I implemented computer vision algorithms to enhance the robot's perception capabilities, enabling it to detect and respond to visual cues essential for navigation and task execution.
• Control System: Developed the control system for the VTOL using MAVLink protocol and Robot Operating System (ROS) to coordinate accurate maneuvers and manage autonomous flight.

Technology Used:

The project utilized OpenCV for vision processing and integrated MAVLink with ROS for communication and control, supporting precise and reliable operation of the VTOL robot.

This project highlights teamwork, advanced robotics programming, and the application of vision and control systems in competitive drone technology.

BrBlind

As a member of the BrBlind team, I was responsible for developing an image processing program designed to read text from books and convert it into audio output through a speaker. This innovative project aims to assist visually impaired individuals by providing an accessible tool for reading printed materials.

Project Responsibilities:

• Image Processing Development: Utilized Python to create algorithms that detect and extract text from images of book pages, enabling the conversion of printed text into a digital format.
• Audio Output Integration: Implemented audio output functionality to play the detected text through speakers, providing an easy-to-use interface for users to access written content.

Target Users:

The tool is specifically developed for blind individuals, empowering them to access literature and educational materials independently.

Technology Used:

The project was developed using Python and ran on a Raspberry Pi, combining powerful image processing capabilities with compact hardware to create a portable reading tool.

For more information about the project, you can read the news article here.

Superball Android Game

I developed a game where the objective is to collect as many items as possible while navigating through various obstacles. Players must demonstrate skill and strategy to achieve high scores and overcome challenges.

Project Details:

• Game Objective: Collect items scattered throughout the game environment while avoiding obstacles that impede progress.
• Gameplay Mechanics: The game features engaging mechanics that require players to strategize their movements and timing to maximize item collection and avoid hazards.

Technology Used:

The game was developed using C++ and the Cocos game engine, providing a robust framework for creating high-performance games with rich graphics.

You can download the game using this link: Download SuperBall.

Make It Android Game

I developed a game where the objective is to make as many cakes as possible according to customer orders.

Project Details:

• Game Objective: Create cakes based on customer orders, ensuring accuracy and speed to maximize the number of completed cakes.
• Gameplay Mechanics: The game challenges players with various recipes and time constraints, requiring quick thinking and skillful execution.

Technology Used:

The game was developed using Java and the libGDX game framework, which provides a powerful and flexible platform for creating 2D games.

You can download the game through this link: Make It.