• Written by Muhammed Abdulla NC, CEO

A quality manager sits at her desk, staring at a spreadsheet of production data, wondering how she can use AI to improve her team's efficiency, but she knows that without the right foundations in place, the project will stall. The business consequence of not having a readiness checklist for quality AI is that it can lead to a significant waste of time and resources, as the team struggles to connect the necessary data and define the key terms. In my experience, this is a mistake that many teams make, and it is wrong to assume that the AI system will somehow magically understand the nuances of the data without proper preparation.

The data connection

The first step in creating a readiness checklist for quality AI is to identify the data that must be connected. This includes not only the production data, but also the maintenance logs, supplier information, and customer feedback. Without this data, the AI system will not be able to provide accurate answers to the team's questions, and the project will break down. I have seen this happen in numerous projects, where the team has tried to launch an AI system without connecting all the necessary data, and it has ended in failure.

A shift supervisor, for example, needs to be able to check the maintenance log and see when the last inspection was done, and what issues were found. This information is critical to ensuring that the production line is running smoothly, and that any issues are caught before they become major problems. Without this data, the shift supervisor would have to spend hours searching for the information, or worse, relying on memory or guesswork.

Agreed definitions

Another critical step in creating a readiness checklist for quality AI is to agree on the definitions of key terms. This includes terms such as "quality", "defect", and "yield". Without a clear understanding of what these terms mean, the AI system will not be able to provide accurate answers, and the team will be working with different assumptions. This can lead to confusion, mistakes, and a breakdown in communication. Most people get wrong the idea that these definitions are obvious, and that everyone understands them in the same way, but this is not the case.

In my experience, it is essential to have these definitions agreed in writing, and to ensure that everyone on the team understands them in the same way. This can be done through a series of meetings and discussions, where the team can come to a consensus on what each term means. It is also important to review and update these definitions regularly, to ensure that they remain relevant and accurate.

Permissions and access

Before launching an AI system, it is essential to decide on the permissions and access levels for each team member. This includes deciding who can see the data, who can ask questions, and who can make changes to the system. Without clear permissions and access levels, the team may struggle to use the system effectively, and may even inadvertently compromise the security of the data. This is a mistake that can have serious consequences, and it is wrong to assume that it will not happen.

For example, a category manager may need to be able to compare two suppliers, but may not need to see the detailed production data. By setting clear permissions and access levels, the team can ensure that each member has the information they need to do their job, without compromising the security of the data. This can be done by setting up different user roles, and assigning each team member to the relevant role.

A first question worth answering

A good first question to ask the AI system is one that the quality team is already waiting on. For example, the team may be waiting on a report on the latest production data, or on an analysis of the defect rate. By asking a question that the team is already waiting on, the AI system can provide a clear and tangible benefit, and can help to build trust and confidence in the system. This can be a powerful way to demonstrate the value of the AI system, and to get the team on board with using it.

For instance, a quality manager can ask the AI system to analyze the production data and identify the root cause of a particular defect. This can help the team to take corrective action, and to improve the overall quality of the production line. By using the AI system to answer a question that the team is already waiting on, the quality manager can show the team the value of the system, and can help to build momentum for its adoption. To get started with this, you can visit our production page to learn more about how our AI system can help with production data analysis.

What this does not do

This readiness checklist for quality AI is not a guarantee of success, and it is not a replacement for careful planning and execution. It is also not a one-time task, but rather an ongoing process that requires regular review and update. Most people get wrong the idea that an AI system can be launched and then left to run on its own, but this is not the case. The AI system requires ongoing maintenance and support, and the team must be willing to invest the time and resources necessary to keep it running smoothly.

If you are considering launching an AI system, but are not sure if you have the necessary foundations in place, you can contact us at contact-us to learn more about how we can help. We can provide guidance and support to help you get started, and to ensure that your AI system is set up for success. By following the steps outlined in this checklist, and by being willing to invest the time and resources necessary, you can create a powerful AI system that helps to improve the quality of your production line.