AI in Manufacturing Shows Promise, But Challenges Lie Ahead

AI promises to be a game-changer within manufacturing, with capabilities such as predictive maintenance and data optimisation. However, there are still come challenges to be ironed out concerning data quality and a lack of understanding over what AI can do.

In considering this, AI Magazine speaks with Herbert Pesch, Managing Director Valtech B2B at Valtech, about the opportunities and risks associated with AI in manufacturing and what businesses within the sector can do to bolster their digital strategies.

He explains that AI promises manufacturers a range of efficiency gains and how it will have one of the most significant impacts on the industry.

What are some of the most promising use cases for AI in the manufacturing sector? 

AI is currently being used to aid predictive maintenance and downtime reduction. Sensors and cameras can provide a wide range of data and continuously monitor individual machine components. This data, in turn, is used to calculate the exact moment a component should be replaced with the least impact on production. 

Valtech's study found that 31% of leaders were already using AI for predictive maintenance and asset optimisation in an established or scalable manner, and we'll likely see this grow as the industry becomes more digitally mature.

Data cleaning and optimisation is another promising area. A third of respondents were found to be using AI to work through the vast datasets held by manufacturers to get them up to scratch for valuable insights. Cleansing and optimisation are tedious and time-consuming jobs however, AI can speed up processing and unlock data for numerous purposes. 

We're also seeing great value in using AI for content creation and product recommendations. There was a massive surge in the availability of AI-generated content in 2023, and increasingly, we're seeing manufacturers leverage this to create more personalised customer experiences and targeted marketing campaigns. This personalised approach to content creation can help boost sales indirectly by enhancing customer engagement, satisfaction, and brand relevance.

What are the biggest challenges manufacturers face when trying to implement AI?

From a data availability standpoint, many manufacturers' operating systems are not set up to make data available for AI applications. Sophisticated AI use requires a solid data foundation, and all too often, valuable data from different business units is not linked or is stored away in disconnected tools. 

In addition, incomplete, outdated and inconsistently formatted data is a real problem and can't be reliably used for AI applications. 

Another considerable challenge is the general attitude towards AI across the industry. By nature, manufacturers are risk averse. While this attitude has stood the test of time for legacy organisations, it's now threatening to become their biggest obstacle. Any introduction of new technology brings concerns around any risks in the production process and a strong awareness of the need to protect trade secrets have come to a head with AI adoption. 

While these attitudes exist for a good reason, leaders need to adopt an open mindset towards the technology and take a measured and educated approach to adoption and implementation. As outlined above, there are plenty of benefits AI can bring to the industry, when implemented in a controlled and responsible manner.  

What organisational or cultural shifts need to happen within manufacturing companies to foster effective AI adoption?

Digital transformation is, first and foremost, a question of organisational and cultural change. 

Setting out a clearly defined AI roadmap is a good place to start as it helps to establish transparent governance structures and foster effective adoption across the entire business. This could include plans to bundle all existing lighthouse projects, which act as beacons for future digital transformation and AI development, into one centrally managed initiative. This will also encourage cross-collaboration between projects to ensure consistency. 

Digital leaders should also consider laying out a strategy for recruiting and retaining specialists, and rolling out an IT setup that enables KPIs to be measured and data to be exchanged quickly.  

Top-down guidance and leadership is required for a change of this scale, so it is crucial to establish C-suite buy-in before embarking on any kind of transformation. Leaders that are well informed on how a data-driven approach leads to new, innovative business models and overall better business outcomes, will be in a much better position to share this knowledge with their teams and ensure this approach trickles down into every level of the business.  This can then act as a catalyst for wider cultural change and inspire employees to be more receptive to, and facilitate, AI adoption. 

What emerging AI use cases do you foresee having the biggest impact on the manufacturing industry? 

Anyone who can predict what will happen in this timeframe, especially in this environment, is looking into a crystal ball. When considering the customer facing domain, such as ecommerce and customer portals, we see a lot of AI coming from the software solutions being used. There is definitely a lot happening in that space, and likely more to come.

We urge manufacturers to get started with AI initiatives, but they should be supported by true business cases. This attitude is exactly why so many manufacturers have been successful for hundreds of years. At the same time, we also see that for many of them, there is still a world to win in other areas of digital transformation. In general, we see a need for a long-term vision, combined with a focus on what brings short term value which is something we are happy to support at Valtech.

For a manufacturing company just starting to explore AI, what initial use case or pilot project would you recommend they pursue first? 

We've found that using large language models (LLMs) to search through documentation can be an easy pilot project, especially in less digitally mature organisations. Documentation can quickly amount to hundreds of pages, making it difficult to access the required information, especially under time pressure. 

By integrating generative AI functions into their customer portals, manufacturers can set themselves apart from the competition and allow clients to get the information they need quickly and with little effort. As an initial use case, the process requires a relatively low lift, with rapid results, which can help kick-start a more comprehensive digital transformation process across the business.

******

Make sure you check out the latest edition of AI Magazine and also sign up to our global conference series - Tech & AI LIVE 2024

******

AI Magazine is a BizClik brand