Industrial companies are increasingly exploring how artificial intelligence can be used to improve packaging efficiency and reduce material waste. According to technology expert Jude Pullen, the shift is not about complex, energy-intensive systems, but rather simple and practical algorithms.
Pullen, who is collaborating with RS Group and Nvidia, explains that the project focuses on optimising packaging through basic computational methods.
– We used simple sorting rules to organise the data better. That accounts for around 80 per cent of the optimisation and requires no advanced AI, he told Packaging Insights.
The system can run on a small device costing around £200, allowing companies to avoid high energy consumption and reduce reliance on external cloud services.
– We don’t need to share sensitive data with large tech companies, he added.
Less waste and fewer errors
The project aims to improve logistics and packing processes by reducing human error and making better use of available space in each parcel. The system checks dimensions, product combinations and packing constraints.
– The goal is to ensure the right quantity of items in the right box while minimising empty space and avoiding issues such as combining heavy and fragile goods, Pullen said.
Reducing unused space is becoming increasingly important as new EU regulations come into force. The Packaging and Packaging Waste Regulation sets a maximum empty space ratio of 50 per cent for grouped, transport and e-commerce packaging from 2030.
According to Pullen, the solution can both reduce material use and lower transport costs.
– The best solutions combine economic and environmental benefits. Using less material means lower costs and fewer shipments, he said.
However, challenges remain. Packaging systems must adapt to a wide variety of products, and standardisation is often limited in real-world operations.
Simple tech instead of heavy AI
Pullen stresses that AI does not have to involve large-scale computing infrastructure. He rejects the idea that companies need to invest heavily in complex models.
– Many assume you need massive computing power. In this case, simple methods go a long way, he said.
He also refers to principles from Japanese industrial practices, where repetitive, difficult or hazardous tasks are automated, but without fully replacing human workers.
– AI should reduce errors, not eliminate jobs, he said.
This balance between automation and human oversight is seen as critical for successful adoption within organisations.
Packaging as a field for innovation
Despite technological advances, Pullen believes packaging still offers significant room for innovation, particularly in accessibility and usability.
– Companies should not see disability as a marginal issue. Solutions that work for more people can also become commercially successful, he said.
He highlights examples such as packaging designed for one-handed use, which can benefit not only people with disabilities but also consumers in everyday situations.
Pullen also encourages companies to explore new perspectives, including regulatory changes, cultural differences and previously overlooked user needs.
– What initially seems like a problem can lead to new and valuable innovations, he said.
Companies can start small, for example, through pilot projects where younger digital talent explores solutions with guidance from experienced team members.
Source: Packaging Insights
Fact check
AI in packaging is mainly used to optimise packing processes, reduce material waste and improve logistics. In many cases, simple algorithms can deliver significant gains without requiring energy-intensive computing systems.
