• Written by Muhammed Abdulla NC, CEO

Every week, research and development teams in large organisations sit through meetings where they discuss failures that have occurred in their programmes, only to realise that similar failures have happened before in other programmes. This phenomenon of failures rediscovered across programmes is a significant problem that can lead to wasted resources and delayed project timelines. The question is, can AI help reduce this problem?

What has to be in the data

To even consider using AI to reduce failures rediscovered across programmes, the data has to be in place. This means that test results and failure analyses, as well as design review records, need to be captured and stored in a way that makes them accessible to the AI system. In practice, this often involves a significant effort to gather and standardise the data from various sources, such as lab reports, testing databases, and design documents. A shift supervisor, for example, may need to work with the quality assurance team to ensure that all relevant data is properly documented and stored.

Most people get wrong the idea that AI can somehow magically extract insights from unstructured or incomplete data. This is simply not the case, and organisations that think otherwise are wrong. The data has to be there, and it has to be in a usable form, for AI to have any chance of helping to reduce failures rediscovered across programmes.

The mechanism

The mechanism by which AI can help reduce failures rediscovered across programmes is straightforward. The AI system is trained on the historical data, including test results and failure analyses, and uses this training to identify patterns and relationships that may not be immediately apparent to human researchers. When a new failure occurs, the AI system can quickly search the historical data to see if similar failures have happened before, and if so, what the root causes were and how they were addressed. This can help researchers to quickly identify potential causes of the failure and take corrective action.

A category manager, for example, may use an AI system to compare the failure modes of different suppliers, and identify which ones are most likely to experience similar failures in the future. This can help the category manager to make more informed decisions about which suppliers to work with, and how to mitigate the risk of failures.

What it will not fix

While AI can be a powerful tool for reducing failures rediscovered across programmes, it is not a silver bullet. There are many reasons why failures may be rediscovered, and AI can only address some of them. For example, if the failure is due to a completely new and unforeseen cause, the AI system may not be able to identify it. Similarly, if the data is incomplete or inaccurate, the AI system may not be able to provide reliable insights.

Furthermore, AI is not a substitute for human judgement and expertise. Researchers and engineers still need to use their expertise and experience to interpret the results of the AI system and make decisions about how to address the failure. AI is simply a tool that can help to augment and support human decision-making, but it is not a replacement for it.

How you would know it worked

So how would you know if the AI system is actually working to reduce failures rediscovered across programmes? One way to measure this is to look at the number of failures that are found before validation. If the AI system is able to identify potential failures early on, this can help to prevent them from occurring in the first place, or at least reduce their impact. A research and development team, for example, may use the AI system to identify potential failures in a new product design, and make changes to the design before it goes into production.

This can be a significant benefit, as it can help to reduce the cost and time required to validate new products or processes. By identifying potential failures early on, organisations can take proactive steps to address them, rather than reacting to them after they have occurred.

When it is the wrong choice

Finally, it is worth considering when AI may not be the right choice for reducing failures rediscovered across programmes. If the organisation does not have a strong data management practice in place, for example, it may not be possible to gather and standardise the data needed to train the AI system. In this case, it may be better to focus on building a strong data management practice first, before attempting to implement an AI system. If you are unsure about how to get started with AI for reducing failures rediscovered across programmes, you can contact us for more information.