AI, Innovation & the Evolving Process of Product Lifecycle Management

At the centre of this transformation is 'Product Lifecycle Management', which has traditionally helped organisations manage information and processes throughout the life of a product, from its initial concept and design through development, manufacturing, maintenance and eventual retirement.

As artificial intelligence becomes more capable, the role of 'Product Lifecycle Management', or PLM, could evolve from a system used for managing product information into an intelligent platform, capable of helping organisations predict problems, automate decisions and identify new opportunities for innovation.

What is Product Lifecycle Management?

PLM is the process of managing a product throughout its entire lifecycle, with a typical product lifecycle involving all or most of the following stages:

1. Concept
2. Design
3. Development
4. Testing
5. Manufacturing
6. Distribution
7. Support
8. Retirement

PLM systems help organisations bring together the large amounts of information generated during these stages, where information may include:

• Product designs
• Engineering drawings
• Technical specifications
• Bills of materials
• Manufacturing information
• Test results
• Supplier information
• Maintenance records
• Product changes
• Regulatory documentation

For complex products, particularly in industries such as aerospace, automotive, engineering, defence and manufacturing, keeping this information accurate and accessible can be a significant challenge, which is where AI could fundamentally change the process and value of PLM.

From Managing Data, to Understanding it

Traditional PLM systems are extremely effective at storing, organising and controlling product information but gaining useful insights within enormous quantities of data, can still require considerable human effort.

AI can potentially analyse this information at a much greater scale, with AI systems already proving that they can be used to examine:

• Previous product designs
• Manufacturing defects
• Test data
• Customer feedback
• Warranty claims
• Maintenance reports
• Supply chain disruptions

Rather than simply recording that a particular component has repeatedly failed, a PLM system using AI, can potentially identify the patterns that would normally remain hidden, flagging potential design or manufacturing concerns within the process, or in the component itself.

This could allow product teams and organisations to move from reacting to problems, towards predicting and preventing them.

AI & the Design Process

One of the most significant areas of development is likely to be the use of AI during product design, where engineers and designers are already using advanced software to model and simulate products.

AI has the potential to take this further by helping generate and evaluate possible design solutions.

An engineer might provide a set of requirements, such as:

• Required dimensions
• Weight limits
• Materials
• Manufacturing constraints
• Performance requirements
• Cost targets

A PLM system being used by engineers and designers, which is assisted using AI, can analyse any requirements and generate potential design options.

This does not necessarily mean that AI will replace engineers, instead, AI could allow engineers to explore a much larger number of possibilities in less time.

The human designer still needs to assess the results, understand the wider context and make the final decisions, so in this sense, AI could become a powerful innovation assistant, rather than an autonomous replacement for human expertise.

Digital Twins & Predictive Innovation

Another important area is the growing use of 'Digital Twins', which are a virtual representation of a physical product, system or process.

For example, an organisation can create a digital representation of an aircraft engine, while using data from the real engine to update its virtual counterpart.

AI can analyse this information to identify unusual behaviour or predict potential problems, which will make stages of PLM, such as maintenance, more proactive.

With the use of AI, instead of waiting for a component to fail, a team or individual engineer, has the potential to identify signs of deterioration and replace the part before a failure occurs.

The same principle could influence future product development, where Information collected from products operating in the real world, can be fed back into the PLM process or environment.

This will involve real world product data being analysed using AI, providing insights to improve the design which will in turn, provide the information needed to replace a component or create the next generation product. Overall, the simple us of AI within the 'improvement', process, will encourage a more continuous cycle of ongoing improvement and innovation.

Faster Product Development

Developing complex products can take years, with teams having to coordinate designs, suppliers, manufacturing processes, testing, compliance and customer requirements.

AI will help reduce some of this complexity by automating routine tasks such as:

• Search technical documentation
• Compare design versions
• Identify inconsistencies
• Analyse engineering changes
• Predict manufacturing problems
• Identify supply-chain risks
• Summarise large technical datasets
• Assist with compliance documentation

This will result in shorter development cycles and faster decision making. However, the speed of development, should not become the primary goal.

In industries where product failure can have serious consequences, such as aerospace, healthcare or transport, analysis and recommendations made using AI, must still be carefully reviewed and validated.

The Rise of Intelligent Product Data

The future use of PLM will depend on how effectively organisations can use their product data, with many companies already possessing decades of valuable information. However, the research and insights they hold, may be stored across multiple legacy systems, databases and documents.

AI can already help organisations make this information easier to access, utilise and manage so an engineer might eventually be able to ask a PLM system:

Have we experienced a similar failure before? or Which suppliers have previously manufactured components with these specifications?

An AI trained model will be able to search across vast quantities of product information and data to provide relevant results, reducing the time spent manually searching for information and allow specialists to focus more on solving problems and developing new ideas.

Transforming Collaboration

Product development rarely involves just one department with a modern product involving engineers, designers, manufacturers, suppliers, software developers, quality specialists, marketing teams and support staff.

PLM already provides a central environment for managing the development of a product across multiple teams, where AI could make collaboration more intelligent by identifying connections between teams and processes.

For example, a proposed engineering change might affect manufacturing costs, supplier contracts, product certification or software compatibility. Using AI will be able to identify any relationships between teams to notify the relevant teams, before any changes are implemented.

This can reduce the risk of changes being made which may result in a possible, negative impact across the whole development lifecycle of a product.

The Challenge of Trust

Despite the potential benefits, the introduction of AI into PLM does introduce significant challenges, especially where product information is commercially sensitive and, in some industries, highly regulated.

Organisations will need to consider:

• Data security
• Intellectual property
• Accuracy
• Bias
• Traceability
• Regulatory compliance
• Human oversight
• Responsibility for AI assisted decisions

One of the most important questions will be whether a company or organisation understands how an AI system reached its recommendation?

For products which are created to be used in high risk environments, for example, medical instruments, a simple answer generated by an AI system will not be enough.

Engineers and regulators may need evidence showing how a decision was reached and what data was used, so the future of AI in PLM is therefore likely to depend heavily on trustworthy and explainable systems.

Replacing the Product Expert

There is growing concern that AI will eventually replace skilled professionals but in product development, the reality is likely to be more complicated.

AI can analyse large quantities of information and identify patterns but experienced engineers and product specialists provide 'Context', which is something that AI is unable to do effectively.

They understand why decisions were made, how products are used and understand risks that may not be visible in the data.

The most effective future model for using AI within PLM may therefore be a mix of human expertise that uses product data with the assistance of AI, to provide faster and better decision making.

AI could take over repetitive tasks and information analysis, allowing human specialists to spend more time on creativity, strategy and complex problem solving.

The Future of PLM

Product Lifecycle Management is likely to become increasingly connected to other technologies, with future PLM environments combining:

• Artificial intelligence
• Machine learning
• Digital twins
• Internet of Things sensors
• Cloud computing
• Advanced simulation
• Robotics
• Predictive analytics
• Generative design

Together, these technologies will create a more connected product lifecycle, with information no longer simply moving from design to manufacturing, to then disappear into the archives. Instead, data from products operating in the real world could continually feed back into the development process.

A New Era of Innovation

The combination of AI and PLM has the potential to change the way products are created.

Instead of relying solely on historical processes and human analysis, companies and development teams could increasingly use intelligent systems to understand their products, predict future problems and explore new possibilities.

The role of PLM may therefore evolve from a system that simply manages the history of a product into one that actively helps to shape its future.

The companies that benefit most may not necessarily be those that simply introduce AI into their existing systems. They may be the organisations that successfully combine AI with product data, experienced people and effective governance.

The Bottom Line

AI is unlikely to replace Product Lifecycle Management but it could fundamentally transform what PLM is capable of doing.

The future may see PLM systems become an intelligent partner in the innovation process, helping organisations move faster from idea to product, while learning continuously from the products they have already created.

As AI continues to develop, the key question for businesses may no longer be whether artificial intelligence will influence product development, but rather, how quickly can organisations build the data, skills and trust needed to use it effectively?