Clarifruit, an artificial intelligence-powered software-as-a-service quality control platform, aims to completely remove subjectivity and human judgment from the produce inspection process, and that goal may not be as far-fetched as it sounds.
In a recent interview with The Packer, Clarifruit CEO Elad Mardix said the platform was created about six years ago and has been commercially available for about three and half years.
The U.S. is its biggest market and accounts for about 40% of the firm’s revenue. With more than 50 customers now, Mardix said the company works with about one-third of the top global produce brands. Most of its early customers have been packer-shippers, but over the last 12 months, after its $12 million A round fundraising, Clarifruit has expanded downstream on the chain into wholesale customers. Mardix said he hopes to announce the first couple of grocery retailers deals in the next few months.
Eliminating quality mismatch
With a customer list that includes Dole, Zespri, Sunkist, Mucci Farms and many others, the company offers quality management solutions that help prevent produce quality mismatches between buyers and sellers across the produce supply chain, Mardix said.
“The mission is to help the industry reduce waste,” he said.
“Our stats from our customers show that about 25% of the transactions in this industry actually fall into the bucket of quality mismatch,” resulting in either a rejection or a price renegotiation, he added.
After speaking with dozens of produce companies in the supply chain, Mardix said Clarifruit found there are three “drivers” that result in quality mismatch.
One reason is that the supply chain doesn’t speak the same language with respect to quality expectations. While every buyer may have a lengthy, paper-based document with its quality specs, those documents are translated in different ways for the seller and the buyer.
Another driver that results in quality mismatches is that the inspection process is human-driven, which can result in inconsistent outcomes.
The third issue is that there is a limited amount of real-time information that is available to help decision-makers.
Instead, decisions are currently based on the intuition of the quality control manager.
“Essentially, what we’ve done is we went to build a software platform that is going to address each one of those issues,” Mardix said.
First, Clarifruit has digitized quality control specs, so any part of the supply chain can download the appropriate quality control specs to their mobile device and perform inspections based exactly on the digital spec as configured by the customer.
In addition, Clarifruit uses AI to enable inspectors in the produce supply chain, whether in the field or at a retail store, to perform quality control with a mobile app.
“Essentially, we developed our own computer vision technology that is able to analyze automatically external attributes,” Mardix said.
Over the last three years, Clarifruit has developed computer vision that can analyze automatically the size and color of fresh produce categories.
“The data goes to the cloud, [and] comes back in two seconds with a grade,” he said.
“What we’ve done here for the grower or a marketing company or a grocery retailer is two things,” Mardix added. “One, we actually improve productivity by about two times, so a 50% to 60% time savings. But the bigger benefit is we’re making the inspection process objective and consistent.”
Another element of what Clarifruit delivers is leveraging the data collected by mobile devices so the supplier can select the appropriate customer for the quality of the produce lot or pallet.
“Do you want to make a decision when you inspect the quality at your [packinghouse] who is the best customer to ship it to?” he said. “We’re the only one in the world that can help these guys essentially either avoid a rejection or maximize the wallet opportunity.”
That solution is delivered through a mobile app for the inspector and a management and analytics dashboard used by the quality assurance management and sales/buyers teams.
External defect solution
Clarifruit was recently one of two companies chosen to develop solutions on driving farm productivity with augmented reality technology in a $9 million investment program through Australia’s Hort Innovation’s Frontiers investment program. According to the Hort Innovation website, the program aims to use the latest augmented reality technology to drive on-farm productivity for berry and table grape growers.
Clarifruit is standardizing the way fruit quality is evaluated to reduce rejections and food waste along the supply chain, the website said. The standardization process will be enabled by fully automating quality control and removing human judgment from the process, according to Hort Innovation. This will be done by providing quality inspectors with AI-powered augmented reality technology along the supply chain, the group said.
“This project will help berry and table grape growers boost their profitability through adopting technology that can improve their fruit picking, automate their quality control processes, improve their training approaches and enhance the way they make on-farm decisions,” Hort Innovation said in a statement.
While this program is focused on berry and table grape growers, there is scope to scale the same solutions to other crops in the future, the website said.
The funding from Hort Innovation in Australia will help Clarifruit accelerate its development of defect automation, which will allow users across the supply chain to automatically identify external defects of grapes and berries, Mardix said.
While Clarifruit’s software so far has focused on size and color attributes of produce, Mardix said the Hort Innovation funding would help Clarifruit reach its goal of automating the entire quality control process.
Identifying external defects is much more technologically challenging than screening for size and color and thus requires many more investment dollars because of costly machine-learning resources, he said.
That has big importance for the produce supply chain, Mardix said. Notably, human judgment will be removed from the inspection process, which should result in a big productivity gain.
“The implication is going to be that the number of people that you’re going to need and the training that you’re going to need to do to a new inspector is going to be much simpler,” he said. “You are still going to need to have a person technician collecting the data, but this person doesn’t need to be an expert anymore, because the algorithm is going to do all the work, the judgment work.”
Unrelated to the Hort Innovation project, Mardix said Clarifruit is finishing the automation of external defect detection for citrus, and that solution is expected to be launched to the market in 2025.
The Clarifruit team will be showcasing the company’s solutions in person at booth No. C1227 during the International Fresh Produce Association’s upcoming Global Produce and Floral Show in Atlanta.
