• Written by Sonali Subhadarsini, Product Manager

When a shift supervisor checks the maintenance log to see if a critical machine has been properly calibrated, they often find that the record of the calibration has not been updated, or worse, the calibration itself has not been done, which can lead to a delay in the production process and potentially affect the quality of the products being manufactured, and this is where AI process automation for quality can help, by automating the steps around the decision to release a batch of products, rather than the decision itself.

Mapping the current process

Before automating any part of the quality process, it is necessary to map out the current process in detail, including all the steps involved, from inspection and testing to certification and release, and this mapping should be done by the people who actually perform the tasks, such as the quality control inspectors and the production supervisors, to ensure that the automation is based on a realistic understanding of the process.

This mapping process can be time-consuming, but it is essential to identify the pain points and the areas where automation can have the most impact, and it should also involve the category manager, who is responsible for sourcing the raw materials and components, to ensure that the automation takes into account the variability in the supply chain.

The goal of this mapping is to create a clear understanding of the current process, including all the handoffs and decision points, and to identify the opportunities for automation, and this understanding will be used to design the automated process, which will be based on the actual workflow, rather than an idealized version of it.

The handoffs that create most of the delay

The handoffs between different teams and departments are often the source of the most significant delays in the quality process, as the information and the responsibility for the product are passed from one team to another, and these handoffs can be automated, using AI process automation for quality, to reduce the delay and improve the efficiency of the process.

For example, when a batch of products is completed, the production supervisor will typically notify the quality control team, who will then inspect and test the products, and if they pass, the quality control team will notify the certification team, who will then certify the products and release them to the customer, and each of these handoffs can be automated, using AI process automation for quality, to reduce the delay and improve the efficiency of the process.

The quality data this depends on

The AI process automation for quality depends on the availability of high-quality data, including inspection records, non-conformance reports, batch genealogy, and customer complaints, and this data should be accurate, complete, and up-to-date, to ensure that the automation is based on a realistic understanding of the process.

The inspection records, for example, should include the results of all the inspections and tests that have been performed on the products, including the pass/fail results, and the non-conformance reports should include the details of any defects or non-conformities that have been found, and the batch genealogy should include the history of each batch of products, including the raw materials and components used, and the customer complaints should include the details of any complaints that have been received from customers.

What it changes about defects and CAPA records

The AI process automation for quality can help to identify defects earlier in the process, by analyzing the inspection records and the non-conformance reports, and it can also help to identify the root cause of the defects, by analyzing the batch genealogy and the customer complaints, and this can help to reduce the number of defects that are traced late to a batch or supplier.

The AI process automation for quality can also help to improve the effectiveness of the Corrective and Preventive Action (CAPA) records, by ensuring that the actions that are specified in the records are actually implemented, and by tracking the results of the actions, to ensure that they are effective, and this can help to reduce the number of CAPA records that are written but never read.

Measured with key performance indicators

The effectiveness of the AI process automation for quality can be measured using key performance indicators, such as first pass yield, defects per unit, and time to root cause, and these indicators can help to identify the areas where the automation is having the most impact, and where further improvements can be made.

For example, the first pass yield can help to identify the percentage of products that are passing the inspection and testing on the first attempt, and the defects per unit can help to identify the number of defects that are being found per unit of product, and the time to root cause can help to identify the time it takes to identify the root cause of a defect, and these indicators can be used to compare the performance of different production lines, or different suppliers, and to identify the areas where improvements can be made, and for more information on production, please visit our production page.

What this does not do

The AI process automation for quality is not a replacement for the human judgment and decision-making that is involved in the quality process, but rather a tool to support and augment the human decision-making, and it is not a solution that can be applied to all quality problems, but rather a solution that should be carefully considered and evaluated to determine if it is the right choice for a particular problem, and for more information on how to get started with AI process automation for quality, please visit our contact us page.