SB'25 San Diego is open for registration! Sign up by January 1st to lock in the pre-launch price!

How Imaging Technology Could Solve the Global Food Waste Problem

One of the biggest food retailers in the US has used the software to better understand the shelf life of its beef, cutting potential waste by 25 percent.

First, some facts. More than a third of all food produced in the world goes to waste at a cost of some $1 trillion. An area the size of China is used to grow this food that is never eaten, not to mention 25 percent of the world’s fresh water supply. Then, there’s the climate change impact. If food waste were a country, it would be the third-biggest emitter of greenhouse gases, after China and the US.

The trouble is, a further 2.3 billion people are likely to join our planet by 2050, demanding a 60 percent increase in the amount of food we need to produce. This just means more waste.

In most developed countries, more than half of all food waste happens at home, at a cost of US$2,275 per family every year in the US.

At the retail level, food waste is around 2 percent of total food waste. However, supermarket practices can often directly affect the amount of food wasted further along the supply chain.

For example, a lot of food wastage happens during product quality control, when samples of food are commonly taken by breaking apart meat, fruit and vegetables to assess their suitability for supermarket shelves. Food that doesn’t pass quality control is either sold for a lower price or thrown away. Workers stand by the side of conveyor belts in distribution centres manually sorting produce or using destructive tests — a hugely inefficient process that only fuels food waste.

This is the specific food-waste problem Abi Ramanan is trying to solve with her company, ImpactVision. After winning a scholarship to Singularity University back in 2015, Ramanan started looking at how sensors, data and software technology might be used to make food supply chains more digital and efficient. Shortly after, she set up her business.

“We discovered hyperspectral imaging, originally developed for use in space, and knew it had great potential as a tool for reducing food waste,” she says. “We decided to embark on the ambitious mission to digitise food supply chains.”

The company’s imaging technology can show the quality of foods — such as the freshness of fish, the tenderness of beef, the ripeness of avocados or the presence of foreign objects. It is non-invasive — meaning no food is wasted in the testing process — and it is quick.

“Removing unnecessary energy use by optimising food utilisation and distribution is a huge opportunity,” Ramanan adds, pointing to research by the Potsdam Institute for Climate Impact Research that shows that 14 percent of agricultural emissions in 2050 could be avoided.

One of the biggest food retailers in the US has used the software to better understand the shelf life of its beef, cutting potential waste by 25 percent. Similarly, the technology can help customers sort avocados based on similar ripeness levels, enabling them to command a 30 percent price premium for delivering consistent products.

The technology can also detect non-magnetic contaminants, such as plastic and paper, that might have found its way into foods. This means that food won’t have to be taken off supermarket shelves and thrown away, something that costs the US food industry $5 billion a year.

Next up, Ramanan says there will be prototype systems for berries, salad and other fresh produce items. And importantly, the team will have built a framework for measuring how much supply chain waste it has prevented and the volume of yield increase being achieved.

In the next 2-3 years, ImpactVision will also launch a smartphone app for assessing fish freshness. And in two years’ time, Ramanan says she hopes to make her API available for third-party developers to build their own solutions using ImpactVision’s data.

“Our products have the potential to revolutionise the way food supply chains operate by providing access to real-time freshness, ripeness or contaminant data for 100 percent of products at production-grade speeds.”