Creating a predictive model to quote accurate procurement costs and lead times for PCBs

Printed Circuit Boards, in short PCBs, are the basis of all electronic equipment, such as computers, mobile phones, and electrical cars. In a PCB, electronic components are connected to another in a controlled matter. Typically, a PCB consists of several components, amongst others electric conductors, metal interconnects, insulators, diodes, and resistors. These components are usually made from a metal such as copper, resin, and glass. The number of elements needed to build a Printed Circuit Board can go up to 100 items. This outlines the complexity of PCB production and the extensive quotation and purchasing process this entails.

Next to the complexity, the scarcity of raw materials – as also reflected in the chip crisis, and supply chain disorders caused by the Covid-19 pandemic, have led to structural supply chain problems and long lead times (e.g., 2 – 3 years) for EMS companies, such as Global Electronics. Therefore, the AMP-subsidised project ColMan between Global Electronics and the FIP-AM@UT was initiated. ColMan is an acronym for Collaborative Manufacturing. This project aims at increasing alternative, flexible solutions in the customer order process, to ensure timely delivery and accurate pricing of PCBs.

From the first stage in the customer order process, a quote request, until procurement and final delivery of the end product, prices and lead times can change a lot. This can for example be due to economic conjuncture, scarcity of materials, or geopolitical issues. A change of prices or lead time between these steps can lead to PCBs being delivered late, financial loss as the difference is absorbed, or cancelled deals.

During the ColMan project we will develop a tool, which predicts the pricing and lead times for the required components at the proposed procurement moment. Taking into account some threshold values, the tool will also be able to suggest suitable alternative materials or components. This proof-of-concept tool collects input from the supply chain and interacts with the customer, to identify, select, negotiate, and offer alternative components selections or designs at an early stage of the customer order process, preferably even before the customer order is made. The goal with this tool is to streamline the customer order process and deliver the best price, lead time, and material bill for the desired product.

Industry partner

APPROACH

The ColMan project proactively tackles PCB production challenges by developing a predictive tool that integrates with the customer order process. This tool provides real-time predictions for procurement costs and lead times, offering greater accuracy and flexibility from the outset. By leveraging supply chain inputs, it anticipates price fluctuations and delays, helping EMS companies adapt to material shortages and geopolitical factors.

Incorporating threshold values, the tool suggests alternative components, allowing quick adjustments without impacting delivery times or costs. Early involvement from customers and suppliers reduces risks from market changes, streamlining procurement and ensuring businesses meet expectations for price, lead time, and material quality while minimizing supply chain disruptions.

OUTCOME

The developed tool accurately forecasts component pricing and lead times during procurement, resulting in decreased costs and shorter lead times. This optimisation enhances customer satisfaction by streamlining the order process, ensuring competitive pricing, timely delivery, and material selection tailored to customer needs.

• Innovative way of automatically replacing risky components using Large Language Models(LLM’s)
• Merging the knowledge from in-house historical order data with the future orders to minimise supply chain problems
• Less dependence on manual checking of incoming quote requests

This cutting-edge solution not only minimizes supply chain disruptions but also enhances the agility and efficiency of procurement, enabling businesses to respond more quickly to market changes while maintaining quality and reliability in their operations. The result is a more resilient, flexible supply chain with enhanced cost control.

This project was made possible through the Regio Deal supported by the Province of Overijssel and the Dutch State.

For more information on ColMan, feel free to reach out to

Engin Topan

Assistant Professor University of Twente

Celal Soyarslan

Celal Soyarslan

Senior Research Engineer

Can Ölmezoğlu

Can Olmezoglu

Engineering Support, Software Development