Clarifruit’s QC revolution

How does Clarifruit’s AI-powered QC software work?

Clarifruit has built the world’s first AI-powered quality control platform exclusively for the fresh produce supply chain. We identified that quality mismatch is the biggest value detractor and the biggest driver of the 45 per cent waste currently taking place in the fruit and vegetable industry. What we’ve done is build an AI-powered software platform that we believe addresses the key drivers for this quality mismatch problem.

Now, how are we doing that? Well, we’ve built a platform that essentially has two components. Component number one is the mobile app, which is available to any inspector, wherever he or she is. She can be in the field, in the packhouse, at the distribution centre, or in the store.

All of those places are already doing sample-based quality control for fresh produce, usually covering about one per cent of the supply. They’re essentially doing it the same way they did in 1950, in an outdated and subjective manner, which is one of the reasons there is so much quality mismatch.

The second component is what we call the management and analytics dashboard ­– the back end of the web app – which is really the brain of the system and allows the QC directors to remotely manage the entire QC process. Most importantly, it provides them with real-time information and insights so they can make decisions based on data rather than what their intuition tells them.

How did you build the Clarifruit platform?

We followed three key steps. Step number one: we invented the Clarifruit language so that you don’t need a 20-page PDF quality spec. For example, a supplier could speak to their apple grower and say: ‘Hey John, before you ship the apples to me, can you Clarifruit the shipment based on a shared digital specification to make sure we’re all speaking the same language?’

Fresh produce suppliers care about the waste, the rejections, and the quality mismatch. If they can expand the sample size and make sure they get the grading correct, so they can ship the produce to the right customer, everyone is aligned and less rejections and waste will result. That’s really what Clarifruit enables them to do. By making the inspection process much quicker and more accurate, suppliers can run a much larger sample size and ensure that they’re not going to have rejections. We officially aspire for above 90 per cent QC accuracy, which is better than any manual, subjective process. For most fruits, we achieve 95 per cent accuracy.

Step number two is our computer vision technology, which we developed in-house. We have a patent in the US and the EU and we’re the only company in the world that’s actually doing it. The technology allows you to take a picture of the fruit or vegetables with your smartphone through our mobile app. It automatically analyses external attributes, such as size, colour or stem colour.

Step three is essentially our big data analytics platform. Currently the big suppliers and retailers may have their own platform, which does a great job of storing the data, but the challenge is that the data is being entered manually by the inspector. It is not collected automatically, so they are very limited in the analytics they can actually run.

With Clarifruit all the data is collected in a structured manner into the mobile app, in real time. This means we can tell an exporter not to ship a certain consignment of fruit to a certain retailer because they’re going to reject it. Then the supplier’s sales guy can decide who they want to ship it to instead. We have the ability to give our customers real-time, actionable insights so they can run their business in a more efficient way.

Until now, the fresh produce industry hasn’t had the capacity to match quality of supply to a specific customer spec in real time; it simply didn’t exist. Instead, suppliers come up with their own specifications. With Clarifruit, we can help these guys optimise the matching of quality to a customer spec to maximise the revenue opportunity and avoid a rejection.

Tell me about Clarifruit’s recently launched automatic defect detection feature and how you think it can benefit the fresh produce industry.

We are also working on automatic defect detection – initially in citrus but I expect that in the next 12-18 months we’re going to develop the feature for all other produce categories as well.

What is the benefit? Reason number one is that by automating QC you’re probably going to see a 50-75 per cent time saving on the actual inspection process. Obviously, it’s much quicker to take a picture of 100 tomatoes or apples at once than it would be to use a size ring, calliper, and colour palette poster to make a subjective judgement on each individual fruit.

Benefit number two, which I think is much more meaningful, is where we make the inspection process objective and consistent. It’s not about what an individual inspector thinks of an apple’s quality, it’s about what the machine thinks, so there should be significantly less quality mismatch between the different parts of the supply chain.

Automating defect identification is going to be a game changer for the industry. Why? Two reasons: firstly, because we can eliminate subjective human judgement from the inspection and grading process, which should allow for a significant reduction in quality mismatch and waste; and secondly, because it will help make the inspection process even quicker. As I said, automating the QC process makes it 50-75 per cent faster, but when you automate defect detection as well, this will rise to a 90-95 time saving.

How can people find out more about Clarifruit?

Easy! Go to our website and book a free demo with our team.