Comment: AI can make a big difference to fruit quality

At Worldwide Fruit, we will soon be able to predict fruit quality and remove subjectivity. That’s because new AI-aided technology and automation are about to make a real difference to our control processes.

We understand how important it is to meet our customers’ specifications. We also know the potential impact if we don’t.

But for every company like ours, a dependence on manual checks makes quality control a bigger challenge. That’s why we need technology, to improve our accuracy.  

Variable quality is extremely disruptive and costly. It leads to inefficiency, loss of yield, food waste, gaps in availability, and worse customer service. But variations are notoriously difficult to measure, and can be extreme.

Climate change especially has led to more frequent, serious weather events which affect crops that are flowering or due to be harvested. Also consider other factors like shipping delays, and it’s true that supply has become less secure, quality more mixed.

So, recently we invested in new technology developed by Clarifruit to automate our quality control processes. It gives us more accurate measurements, removes the subjectivity, and gives us confidence in our decision-making.

As a business, we need trusted, accurate data so we can make the right decisions with our suppliers and customers. We must identify problems accurately and early. Then the right quality fruit flows through our storage, ripening and packing facilities as efficiently as possible.

As an industry, we must control better, and achieve greater levels of ‘right first time’ quality across our end-to-end value chains.

To do this, we must automate and digitalise assessments where we can, including further back along the chain where fruit is packed at source. No more manual size or colour checks, no more human bias and error. Instead, we can embrace digital measurement of external defects; expand our sample sizes to check more quickly and accurately; and use cameras and AI as Clarifruit does to predict shelf-life.

After this, our assurance process can be risk-based, because our own suppliers will generate quality data that can be verified. Fruit that doesn’t meet our customers’ specifications will be downgraded and redistributed earlier. Claims will be minimised and grower returns maximised.

In short, this exciting step into AI-enhanced technology should mean we continue to deliver the best-quality products and the best service for decades to come.