News Article

FMEA Speeds Time To Market In Photonic IC Manufacturing

Failure mode and effects analysis provides a framework for rapid design and manufacturing implementation that has contributed to the successful launch of a new product suite by MetroPhotonics, according to Steven Martin and Bedwyr Humphreys.
Failure mode and effects analysis (or FMEA) is a simple yet powerful pro-active engineering quality method that is designed to help identify and counter weak points in the early conception phase of products and processes. The structured approach makes it easy to use, and the benefits far outweigh the investment in time and resources needed to execute the analysis. Put simply, FMEA is a rigorous risk-analysis tool that is used during the product concept phase to evaluate high-risk stages, serving to focus and define development effort.

FMEA is fast gaining favor across a number of industries, including the silicon IC industry. Its application to the compound semiconductor industry yet again reinforces the industry trend of shifting away from laboratory practices towards well-defined large-scale manufacturing methods, as the industry matures and cost becomes an important driver.

Added value

The added value of the FMEA process, as compared with other quality techniques, is that it is a continuous process and not an event. Corrective action requests (i.e. customer issues) are resolved, continuous improvement initiatives are identified, and future advanced designs and the product Control Plan are evolved. It is a disciplined technique that can help companies committed to continuous improvement (i.e. lean manufacturers) identify and eliminate potential waste.

Broadly speaking, there are three types of FMEA, namely Concept FMEA, Design FMEA and Process FMEA. Each represents a stage in the product development cycle where one follows the other as the new product is brought into production.

Concept FMEA (C-FMEA) is used to optimize system alternatives, which helps formulate a customer specification. In terms of being applied at MetroPhotonics, many of the system alternatives had already been established and it was more important to implement a product rather than review a well-established path. However, the benefits of a C-FMEA are as follows.

• Generates failure-mode occurrence ratings that can be used to estimate whether a particular concept alternative can achieve its reliability target.
• Increases the likelihood that all potential effects of the concept proposal failure mode will be considered.
• Identifies potential failure modes caused by interactions within the concept and helps to determine if hardware redundancy is required within a design proposal.
• Forms the basis for developing system-level diagnostics procedures.
• Forms the basis for developing system-fault management techniques and helps determine changes to system-design specifications.

Based on the specifications defined in the C-FMEA, the Design FMEA (D-FMEA) helps in the objective evaluation of design requirements and alternatives for manufacturing and assembly needs. In this industry it helps with the planning of thorough and efficient design, test and development programs. At completion, the D-FMEA prioritizes design improvements and develops testing actions that can be managed in a project management format.

Based on the D-FMEA, the Process FMEA (P-FMEA) assists in the analysis of new manufacturing and assembly processes. It identifies process deficiencies, thus enabling engineers to focus on controls for reducing the occurrence of, or increasing the detection of, unacceptable products. On completion, the P-FMEA establishes a priority for process improvement actions and hence another round of project management.

The timing for using these tools is as follows. A C-FMEA is started as early in the program as possible, to establish the specifications. A D-FMEA is initialized before the design concept is finalized, and completed before the production drawings of parts are released. Finally, the P-FMEA is started before tooling for a production process or the feasibility state when all processes and parts are being considered for production.

FMEA analysis

The basis for FMEA analysis is to categorize failure modes based on their severity, occurrence and detection. Three tables are generated that grade each of these characteristics on a scale of 1 to 10. Each table can be customized, and so specific product requirements are captured, allowing for tailored analysis. For example, in the automotive industry safety is critical and would be graded a 10 in terms of severity given a particular failure mode. In the photonics industry, a severity of 10 would be attributed to the complete failure of a product in the field, shutting down a customer s major systems with no (redundancy) backup. Occurrence and detection tables do not normally need to be modified unless it is very difficult to establish a ranking where all failure modes fall within the same occurrence or detection rating.

A typical FMEA form (spreadsheet) lists all of the critical requirements for the FMEA under the following headings.
• Item/Function: the functional part of the product design to be analyzed.
• Potential failure mode: the manner in which the function could potentially fail to meet design intent (this is determined under group brainstorming).
• Potential effects of failure: the effects of the failure mode on the function as perceived by the customer.
• Severity: ranked from 1 to 10,1-2 indicates a failure so minor that the customer will probably not detect it, 10 indicates a failure that will result in major customer dissatisfaction and cause system non-operation or non-compliance.
• Classification: used to classify failure modes other than by severity, occurrence and detection. This is because of special government regulations, customer satisfaction, safety etc, and these are identified as either critical characteristics or significant characteristics.
• Potential cause: describes the design weakness that causes the failure mode. The D-FMEA does not rely on process controls to overcome weaknesses.
• Occurrence: ranked from 1 to 10, 1 indicates a very low probability of occurrence, 10 indicates a high probability of occurrence, i.e. once a week.
• Current design controls: methods or tests used to detect a first level cause of a failure mode, or the failure mode itself.
• Detection: ranked from 1 to 10, 1-2 indicates that verification and/or controls will almost certainly detect the existence of a deficiency or defect, 10 indicates a very low (or zero) probability that the defect will be detected.
• Risk priority number (RPN): product of the severity, occurrence and detection rankings, in which the potential design weaknesses are ranked for possible redesign actions.
• Recommended actions: actions that will reduce the severity, occurrence or detection of a particular failure mode so as to reduce the size of the RPN.
• Responsibility and target completion dates.
• Actions taken, severity, occurrence, detection and RPN: once actions have been taken, the analysis is reconducted and a new RPN derived.

As the information flows from the C-FMEA through the D-FMEA to the P-FMEA and the Control Plan, the level of data increases exponentially. One item in the C-FMEA may raise three items in the D-FMEA, each of which raise a further three in the P-FMEA, which ultimately raises another three for control (see figure 1). This may appear to make things more complicated, but facing up to potential problems as early as possible in the development is essential, providing a means to more accurately determine realistic timelines and project costs.

To determine which item is most in need of attention, the severity, occurrence and detection values are multiplied together. The item with the highest score has the highest risk. In the example shown in the process table above, the main focus for process development would be to improve the score for the photoresist dispensing process. This could be done by either improving detection (reducing the score), by measuring every product wafer prior to exposure (not a manufacturable solution), or by reducing the occurrence through improved hardware control (spin-speed repeatability) and greater preventative maintenance frequency, in order to eliminate uncontrolled process variability.

CS International to return to Brussels – bigger and better than ever!

The leading global compound semiconductor conference and exhibition will once again bring together key players from across the value chain for two-days of strategic technical sessions, dynamic talks and unrivalled networking opportunities.

Join us face-to-face between 28th – 29th June 2022

  • View the agenda.
  • 3 for the price of 1. Register your place and gain complementary access to TWO FURTHER industry leading conferences: PIC International and SSI International.
  • Email  or call +44 (0)24 7671 8970 for more details.

*90% of exhibition space has gone - book your booth before it’s too late!


Search the news archive

To close this popup you can press escape or click the close icon.
Register - Step 1

You may choose to subscribe to the Compound Semiconductor Magazine, the Compound Semiconductor Newsletter, or both. You may also request additional information if required, before submitting your application.

Please subscribe me to:


You chose the industry type of "Other"

Please enter the industry that you work in:
Please enter the industry that you work in:
Live Event