Innovating at the Intersection of Automation and AI
Your journey towards innovative solutions starts here.
Explore Featured ProjectsFeatured Projects
Digital Inventory Lending System
Independently conceptualized and developed a digital inventory and lending system, designed to streamline the management of university event resources. This web application, powered by a robust SQLite database, facilitates efficient tracking and lending of equipment, ensuring availability and preventing double bookings.
Technologies
Outcomes:
- User-friendly GUI Web-App, making the system accessible to all university members
- Significantly reduces processing time for inventory requests.
- Enhances booking reliability and avoids double bookings.
- Facilitates better planning and lending of inventory items.
- The project is still under development, with plans to deploy the web application on a production server for broader access and use.
Project Reflections:
- Initiated the project with a TKinter-based prototype to establish and refine the backend logic, demonstrating the value of iterative development and early validation of concepts.
- Implemented sophisticated time logic for reservations, including real-time updates to article availability, ensuring the system reflects current inventory status accurately.
- Integrated automated email notifications to confirm reservations with clients, enhancing communication and user experience.
- Optimized SQL queries and employed indexing strategies to speed up data retrieval processes, significantly improving system performance and user satisfaction.
- Learnt the importance of user feedback in the early stages to streamline the UI/UX design.
- Improved my skills in Python, Flask, and database management through hands-on application development.
AI-Based Chip Detection System
Led the development of an AI-based chip detection system for turning machines using cutting-edge YOLOv8. This project encompassed the full spectrum of system development, from initial data acquisition and preparation to the final stages of model training and inference.
Technologies
Outcomes:
- Developed a user-friendly display device based on Raspberry Pi for seamless interaction between machine operators and the AI inference, improving operational efficiency and user experience.
- Achieved high accuracy in error classification and detection with the integration of the YOLOv8 object detection model.
- Implemented a powerful, Linux-based AI computing server to support real-time AI inference, enhancing the system's integration into the existing IoT network.
- Designed and implemented a mobile data acquisition station, optimizing AI training data collection under varied environmental conditions.
- Established a foundation for scalable AI applications in manufacturing, paving the way for further innovation and efficiency improvements.
Project Reflections:
- Crucially, the project involved the iterative development, testing, and optimization of prototypes, ensuring each version improved upon the last until successful implementation was achieved within the production environment. This process not only demonstrated the prototype's viability but also its adaptability and scalability in real-world settings.
- Effective project management and documentation, including the use of GitLab and comprehensive technical manuals, were critical for maintaining project momentum and fostering team collaboration.
- The project highlighted the necessity of a solid data foundation, emphasizing data acquisition and preparation as key steps in developing reliable AI models.
AI-based air detection system in magnetic metering pumps.
Developed a deep learning model to classify errors in time series data, involving the creation and preparation of the dataset for effective model training. Successfully deployed a TensorFlow Lite version of the best-performing model, optimized for TinyML applications. This deployment enables efficient, real-time error detection in resource-constrained environments, significantly enhancing predictive maintenance capabilities..
Technologies
Outcomes:
Project Reflections:
Professional Portfolio Development
Created my(this) web portfolio to share my profile and a range of projects.
Technologies
Outcomes:
Project Reflections: