The scale of our modern food supply chain is dizzying: 5,000,000 different products from trillions of tonnes of harvests, changing hands an average of nine times before it lands on our plates. For the most part, the system operates as a "push" market, where producers push products with very limited visibility into demand, resulting in a vicious cycle of overproduction and waste, as well as an almost total lack of accountability at every level. Farmers sell to harvest managers, harvest managers sell to brokers and so on, with each layer abdicating responsibility once the product has been pushed to the next step.
The end result is that 40% of all food produced worldwide goes to waste, with some estimates putting this figure as high as 50 to 60%. This means we plant, irrigate, fertilise, transport, refrigerate, package and ship twice as much food as we actually need, ultimately generating massive amounts of carbon emissions that make food waste the second-largest contributor to climate change behind energy.
As the co-founder and CEO of a company whose long-term mission is to end food waste, this is an issue that I am particularly passionate about – and frankly, one that keeps me up at night. This year, however, I am feeling more optimistic than ever about our ability to solve this pressing issue because of one thing: AI.
The potential of AI
Across the board, the food industry has one of the lowest levels of digitalisation of any sector.The lack of digital tooling and data-driven decision making, and the amount of work and change management required to adopt new technology means that things move slowly and the status quo remains unchanged for years at a time. With the arrival of AI, we have the potential to change all of this. In fact, it’s hard to overstate how transformative this technology could be at every level of our food supply chain.
Farmers and manufacturers could tailor their output and use more sustainable growing methods. Every product could be tracked across its journey, from source to plate. That journey could be optimised, with inefficiencies tracked and recommendations made on how to cut waste, emissions, and overheads. Non-value-adding intermediaries could be omitted, and food could travel the shortest route possible to reach its destination. We could transition from a push market to a waste-free pull market where supply and demand are balanced.
This is a bold vision for the future, of course, but there are powerful technologies available today that, if adopted, can help us reach it. In the near-term, I see five specific areas with the greatest potential for disruption and improvement by AI.
Food quality. Between refrigeration issues, incorrect storage, food expiration and contamination of transport containers, food quality is a concern that can lead to waste at nearly every step of the supply chain. Companies like Amazon are already using AI sensors to monitor storage and transportation conditions in real-time to proactively prevent waste and even predict shelf life when food is on its way to its destination.
Quality control. In produce, where waste is an especially big problem, AI can also be used to solve for waste that results when perfectly good produce is thrown away because it doesn’t meet arbitrary cosmetic standards. By replacing subjective, manual quality control processes with digital, data-based inspections, we can significantly cut down on the more than 20 billion pounds of so-called “ugly” or unharvested food that are thrown away each year.
Demand forecasting. Until recently, manual data entry and tedious analysis from someone with specialised skills would have been needed to make even the most primitive of demand forecasts. Today, AI can read delivery notes, review sales data, consult weather forecasts and interpret trends in order to tell a distributor exactly how many tomatoes will be needed in London in March of next year. And that person doesn’t need a degree in computer science or even extensive training – they can simply type the question into an interface. This kind of data is already being used downstream to make buying decisions, but the hope is that it will eventually help guide us to a place where we aren’t growing more than we need.
Optimized inventory and order management. Manual, spreadsheet-based processes for ordering and inventory management have historically led to human error and over-ordering that leads to waste. Today, new AI-driven inventory management systems and order management tools allow restaurants and distributors to better monitor their stock levels to know when replenishments are needed, so they are ordering the right amount of food at the right time, as well as streamline the order process to avoid mistakes and duplicate orders.
Food redistribution. New technology can help restaurants avoid sending surplus food to landfill, either by matching excess food from restaurants and other hospitality businesses with local non-profits, or by allowing consumers to buy leftover, expiring, or misshapen food at discounted prices from local eateries.
Looking towards the future
We are already facing unprecedented challenges as a result of climate change, and our food system will need to feed a population of 10 billion by 2050. This challenge, coupled with the inefficiencies in our current food system, means that we’re rapidly moving towards a future where our food system wreaks havoc on poverty, equality, and even political stability, making this an urgent humanitarian need.
My hope is that spreading awareness of this issue will encourage individuals and businesses to take action –including the adoption of AI – at every level of the food supply chain. Whether you are one of the 1.23 billion people who work in the food industry, or you are simply a consumer, we all have a role to play in cutting down food waste. Let's seize this opportunity for lasting change, starting today.