AI & Packaging: Just get started!

neue verpackung: Mr Carstens, how does AI influence your daily work?

Moritz Carstens: AI has become an integral part of our day-to-day work, and we now use it as standard in various project phases. For example, trying out different styles is more efficient and multifaceted thanks to generative AI. AI also helps with routine tasks such as visualising prototypes and conducting market research. We also use AI to simulate small-scale market research. Data protection is no longer a significant issue either. We have developed our own secure tools, ‘Ways of Working’ and a ‘Code of Conduct’ that enable us to safely work with these new tools on behalf of our customers.

neue verpackung: How has the understanding and acceptance of AI developed in your teams? Or among your customers?

Moritz Carstens: We believe in the interplay between humans and machines, combining ‘human intelligence’ with ‘artificial intelligence’. This means that designers become initiators and curators. The machine takes over elaboration and reproduction at various touchpoints. The resulting brand intelligence helps our customers increase the quality and efficiency of their brand management.

AI is a powerful tool, but it is not intelligent without humans.

neue verpackung: Which new AI technologies or tools do you consider particularly relevant for packaging development?

Moritz Carstens: Generative AI models are currently highly relevant. This is true in both the image and text areas. We use generative image AI, such as Firefly or Midjourney, to quickly generate visual ideas in different styles for various touchpoints, a process that would previously have taken days of illustration or photography. I also find voice models exciting. With the right model, we can adapt tonalities specifically or create translations. This works impressively well.

We are also seeing the emergence of invaluable AI-supported analysis tools for packaging designers.

Specialised packaging AI platforms are also becoming increasingly popular. Some providers can already generate print-ready PDFs or InDesign files at the touch of a button, provided a scalable design system is in place.

neue verpackung: Have you developed any specific strategies for either data collection or networking?

Moritz Carstens: Over the last two years, we have invested heavily in our data strategy. We primarily use data that we have collected and curated ourselves to feed our AI tools, or we use models based on licence-free material. Networking data is important to us. Silos are not helpful for AI. That’s why we try to link different data sources, such as design assets with consumer feedback or production data. For example: We analyse the effect of certain design decisions, such as colour or material choices, on sales figures or consumer behaviour. This type of networked data provides AI with context, enabling it to make more informed recommendations.

neue verpackung: What are the biggest hurdles to the use of AI in packaging development right now?

Moritz Carstens: Technically, generative AI is still limited in the packaging sector. While we can create impressive renderings and designs using AI, converting them into print-ready files is time-consuming. AI-generated layouts usually have to be cleanly transferred into our graphics software by designers. Without human fine-tuning, this is not yet possible.

Another hurdle is the procurement and release of data: companies often have valuable dormant data (for example, from market research or production quality) that would be useful for AI models. However, this data is held in different departments or is difficult to utilise for data protection reasons. This silo mentality needs to be broken down.

neue nerpackung: What developments are expected in the next 12 to 24 months?

Moritz Carstens: Many software providers, including Adobe and specialised packaging software companies, are already working on seamlessly embedding AI functions into design environments. This means that users may not even realise they are using AI when layouts are automatically adjusted or image content is generated with the click of a mouse, for example.

I also expect to see considerable progress in multimodal AI systems. These systems can understand and generate images and text simultaneously. For us, this could mean an AI analysing a packaging image and immediately providing suitable improvement suggestions or marketing texts — 360-degree support, so to speak.

The link between AI and intelligent packaging has received relatively little attention, but it is a very interesting topic. I’m thinking of packaging that uses sensor technology or digital printing to deliver customised messages — here, AI could generate real-time content tailored to users or situations.

neue nerpackung: What advice would you give to companies that are planning to start using AI for packaging development? What should they look out for, and where should they start?

My most important piece of advice is to start simply, but with a clear focus. First, companies should identify where AI can have the greatest impact on their packaging process.
They should also consider the framework conditions early on and develop a company-wide attitude. Many of our customers are now examining their brand essence to derive an AI mindset from it. For example, if your focus is on real people, it may be unwise to depict people using AI, as this could contradict your ethics. This has to be defined individually: How and where do we want to use AI, and above all, why?

This is an extract from an interview with Philipp Bittermann from ‘neue verpackung’.