Faberwork helped automate the semiconductor pipeline process.
We show how data analytic integration can boost productivity and avoid bottlenecks in printed circuit board (PCB) production.
Client
Our client was a manufacturer of customized PCBs. The Client accepted PCB design orders via the web. They offered the PCB design with available components in inventory. Once a client approved the design, the PCB fabrication was done.
Challenge
The Client faced challenges in optimizing their PCB layout process. With a commitment to innovation and efficiency, they sought a solution that would streamline their workflow, reduce production costs, and enhance the overall product quality. 
The Client's existing PCB layout process was time-consuming and prone to errors. Manual placement of components often led to suboptimal designs, resulting in increased production time and material waste. They needed a sophisticated solution that could automate layout optimization while ensuring compatibility with their existing systems.
Some of the issues included:
- Board size constraints
- Mechanical component integration
- Use of inventoried items only
- Board thermal considerations
- Utilizing power efficiency
Regarding inventory management, if a customer were asking for a certain type of transistor and it was not available in inventory, then the Faberwork model would suggest layout changes for substitute components. The Faberwork model could also recommend layout changes for a new component size, integration constraints, or thermal needs.
Solution
Partnering with Faberwork, the Client had a transformative journey to improve their PCB layout process. Leveraging the power of Python for backend development, Flask for web framework, Eve for RESTful API, and Angular for frontend development, we created a comprehensive solution tailored to their unique needs.
Backend Development with Python and Flask:
Utilizing Python's versatility and Flask's simplicity, we built a robust backend infrastructure to handle complex optimization algorithms. This allowed seamless integration with the Client's existing systems and ensured scalability for future enhancements.
RESTful API Development with Eve:
Implementing RESTful APIs using Eve facilitated efficient communication between frontend and backend components. This modular approach enabled flexible data exchange and simplified integration with third-party services, enhancing the solution's versatility.
- Database schema: We built a robust database schema where we stored all the generic and brand components.
- Searching components: We stored all of the available components in the database. If a specific brand was not available in inventory, we identified the stock level as zero. This helped us identify similar components. API is created in Python Eve to provide the best alternative components. Faberwork created an algorithm that helped them choose the best fit.
Front-End Development with Angular:
Employing Angular for front-end development, we crafted an intuitive user interface (UI) that prioritized user experience (UX). Interactive features, real-time updates, and customizable layouts empowered users to visualize and manipulate PCB designs with ease, fostering efficiency and creativity.
- Re-Layout Attributes: Besides storing the technical specifications, we also stored these attributes:
- Footprint size
- Tolerance
- Drift in PPM/deg C
- Voltage rating
- Thermal needs
- Re-Layout of PCB: If the existing PCB design needed any updates as a result of extra Re-Layout parameters, then the model showed the new design using the Angular MVC framework. If this new layout had any size, thermal, or power issues, then a new layout was produced with a new component. In this way, Faberwork’s model helped the customers' design engineers choose another component, which was available in inventory.
PCB Layout Optimization Algorithms: Leveraging Python's extensive libraries, we implemented advanced optimization algorithms to automate the PCB layout process. From component placement and routing to signal integrity analysis, Faberwork’s solution ensured optimal performance while minimizing production costs and material waste.
- Self-organizing genetic algorithm (SOGA): We used a self-organizing genetic algorithm (SOGA) to reorganize the components with alternatives. In this way, a re-layout of the PCB was quickly achieved. Design engineers were able to accept or reject designs with the help of warnings and issues from the model.
Results
The implementation of our PCB layout optimization solution yielded strong results for our client:
- Increased Efficiency: Automated layout optimization reduced design iteration time and minimized manual intervention, leading to significant time savings and increased productivity.
- Cost Reduction: By optimizing component placement and routing, our solution minimized material waste and reduced production costs, resulting in substantial savings for the Client.
- Enhanced Product Quality: Optimized PCB layouts improved signal integrity, reduced electromagnetic interference, and enhanced overall product quality, leading to greater customer satisfaction and loyalty.
- Scalability and Flexibility: The modular architecture of our solution, coupled with seamless integration capabilities, ensured scalability and adaptability to evolving business needs. This future-proofed the Client's operations.
“Through the seamless integration of Python, Flask, Eve, Angular, and SOGA, we successfully transformed our Client's PCB layout process. This increased efficiency, cost savings, and product quality.”
—Yogesh Sharma, Senior Vice President (Delivery) at Faberwork
