As we approach 9 billion mouths to feed globally by 2050, there is a need for a dramatic leap in transitioning to food systems that are sustainable. Artificial intelligence (AI) could become the key to that transition, but it needs to be done with humans at the centre of AI design. A recent academic paper published in the Journal of Cleaner Production by Food Agility CRC and RMIT PhD student Stéphanie Camaréna, investigated the intersection between AI, sustainable food systems and design.
The ability to secure the food and nutritional needs for current and future generations is facing extreme challenges. Population growth, climate change, rapid depletion of resources, freshwater depletion, and pollution of environments are all pointing towards a food security crisis. And agriculture and food production systems right through the supply chain are a major contributor to these issues.
Food systems contribute significantly to climate change, land clearing, high use of fertilisers, emissions from livestock and the storage, transport, and processing of food. At the retail and consumption level there is significant amounts of food waste, which if food waste was a country would be the third largest greenhouse gas emitter in the world.
A fundamental change in how food systems function will be essential to avoid the collapse of food security and nutrition. This is where AI comes in as a solution to a sustainability transition.
AI is less a futuristic novelty and now a current day technology. Already AI is being used in agriculture and food production in ways that have the potential to transform food systems for the better, however, there are many considerations to ensure this is done as ethically as possible with a human-centred approach.
AI is and will continue to have an impact on how we produce, supply, and consume food. AI in food systems can be applied several ways: through the management of crop health (computer vision, machine learning, pattern recognition, robotics), the automation of farming operations (robotics) and the development of demand-driven supply chains (pattern recognition and prediction, natural language processing).
Tracing food from paddock to plate is becoming a consumer demand, and AI can bring that level of transparency to the consumer and retailer. Food waste can also be reduced by using food recognition algorithms to identify food stored in the freezer and using intelligent packaging.
Precision health can drive healthy eating habits and informed consumer choice with AI-powered tools, learning from consumers’ food habits to target recommendations on healthy food choices, diet, and food safety.
“AI will be able to reduce the food system footprint by 20-30 percent which is great. But while AI is an enabler of sustainability, it also has a cost,” said Stephanie.
“It has impacts that aren’t so positive, with the training of a complex AI algorithm producing as much in the way of emissions as flying from New York to San Francisco 300 times.”
Designers are the mediators between people, technology and the environment. Stephanie is a design researcher and she acknowledges the critical role that designers should play in designing sustainable food systems alongside AI.
“AI can 'think' in 100s of dimensions, it is a powerful tool. It can do what humans might find difficult to think about. But AI needs to be driven by experts, with the human aspect guiding what we want it to do.
To date, I have found a lot of tech driven solutions that are presented to food producers and they are not practical, so they fail to get take-up. The solution is co-design with everyone involved and starting with what is the problem, not what is the technology.”
There are pros and cons to AI technologies, dependent on how they are applied. Stephanie’s research found that while AI can contribute to environmental benefits, such as decreased water and oil use, there can be trade-offs over ethics, fairness, access and transparency.
“AI is often presented as a technological ‘saviour’ that can solve the problem of sustainability. We must put sustainability as the main aim in AI or it just becomes an afterthought.
To create human-centred sustainable food system solutions, they need to be co-created and include a cross-disciplinary approach that also involves government and communities. This type of design thinking helps balance AI with the needs of the environment and of people. We have an opportunity as designers to look at how we integrate and engage with AI, and with the food producer and consumer,” she said.
Stephanie will be expanding on this topic with further research publications, including on co-design and AI in farmers markets.
Stéphanie Camaréna, Artificial intelligence in the design of the transitions to sustainable food systems, Journal of Cleaner Production, Volume 271, 2020, https://doi.org/10.1016/j.jclepro.2020.122574
Student profile: Stephanie Camarena
Host University: RMIT University
Academic supervisors: Dr Judith Glover, Assoc. Prof. Karli Verghese