AI-powered packaging innovation uses sensors to reduce food waste

Researchers from Kyushu University, Japan, have developed a framework for “future-ready food packaging” designed around a closed loop of recognition, judgment, actuation, and feedback. The innovation aims to detect what’s happening inside the packaging in real time and translate that into information for producers and consumers. 

The study, published in Trends in Food Science & Technology, combines three research areas said to have developed in isolation — intelligent sensing, self-healing materials, and AI-driven prediction — into a single, coherent system.

Fumihiko Tanaka, professor at Kyushu University’s Faculty of Agriculture, says: “Globally, roughly one-third of all food produced is wasted. It also carries a climate cost, as food loss accounts for roughly 8% of global greenhouse gas emissions, approaching the roughly 10% attributed to road transport.”

The researchers note that food waste often occurs during distribution due to damage, as well as through unnecessary disposal. Products are frequently discarded before spoilage due to inventory turnover pressures or printed date labels rather than actual food quality. 

Confronting the reality, Fanze Meng, the paper’s first author and a postdoctoral researcher at Kyushu University, details: “Future‑ready packaging requires a different mindset. We wanted the film to communicate with the food itself, converting optical or gas signals into electrical data, and using AI to interpret what’s happening inside.”

Quality monitoring

After a systematic review of technologies, the researchers placed sensors in the packaging, detecting the pH shifts, gases, and microbial byproducts that signal spoilage. 

Natural pigments such as anthocyanins, the compounds found in foods like purple sweet potatoes, can shift color as pH changes, providing a readable signal at every stage of spoilage. For example, in spoiling meat, alkaline gases accumulate, and the material shifts continuously from purple-red to yellow-green.

Xirui Yan, a researcher for the Japan Society for the Promotion of Science at Kyushu University, explains: “To survive real‑world distribution, the material needs more than just the ability to sense. Light and heat can cause false readings, and a bump or scratch can interrupt the signal, so reliability must be engineered in.“

”One approach we’ve tried is anchoring the pigments with metal-organic frameworks and carbon quantum dots and adding self-healing capacity, so the film keeps working even after damage.”

Next, the proposed AI-enabled system converts spoilage signals into responses, such as releasing antimicrobials, sending alerts, or optimizing logistics. “It’s like giving produce a full check-up. The film collects the signal, AI analyzes it, and together they tell you the food’s condition and what to do next,” says Yan.

Data interpretation 

The research team plans to update the recommendation system to reflect how differently foods spoil. By tracking the compounds each food releases as it spoils, the film captures different patterns that AI can learn from. According to the researchers, the data gathered by the sensors could assist material designers and food producers in adapting solutions to different foods.

“A deeper understanding of how produce deteriorates can also inform sales and consumption strategy,” says Tanaka. 

The research team is collaborating with local governments and logistics partners to explore methods to grade produce based on how well it withstands storage and transport. To reduce food waste, short-shelf-life items would go to local markets, while hardier varieties are reserved for export.

At the consumer end, the solution could be delivered through connective packaging, enabling consumers to receive instant information about the food product via scanning, notes the research.

The researchers share that future studies should look into the long-term safety and stability assessments for food-contact materials, particularly certain nanomaterials. Moreover, consistent quality control at an industrial scale requires further research.

“We are setting a direction. We hope others will build better ones on top of ours. If enough people move together, it becomes a beam of light, then a path — and eventually, that path could lead from the lab to something real,” Meng and Yan conclude.