Automated quality checking is a more precise and time efficient way to carry out a task that is rejecting huge volumes of perfectly good produce before it even reaches retailers. Researcher Rick van de Zedde explains, in a preview of his seminar on Phenomenal Phenomics – new ways to determine quality at The Amsterdam Produce Show and Conference on November 3
When there is talk of robotics taking on human tasks within the fresh produce industry, often the tone is that of future developments. However, that technology is here, and is already being deployed in some cases, or is ready to roll out with further investment.
Rick van de Zedde, a senior researcher and business developer for computer vision at Wageningen University & Research Food & Biobased Research, is in the latter situation with an objective robotic system for checking quality and decay of fresh produce.
Traditionally humans are charged with assessing the quality, and subsequent decline during the supply chain, of fresh produce. Whether it is immediately post-harvest, or on arrival at its intended destination, huge volumes are judged on just a handful of samples.
Due to time and labour constraints, the assessor can only check a few products, using sight, smell, taste, and a small set of sensors for instance, to measure the brix level of a fruit. It is then up to them to determine whether the whole shipment is according to the expected quality and right for consumption.
This presents a multitude of potential problems for growers, trading companies and retailers, as van de Zedde explains: “Not only is it quite a dull and repetitive job, experts do not always agree on quality, and between different companies there can be disagreements, so the testing is not always consistent.
“The impact of this is huge; there is a large monetary loss should a company reject batches of produce, and when you don’t have uniformity within those batches, then it could come down to luck. If you’re taking 10 samples out of thousands of fruit, that cannot represent the whole shipment.
“This is where we come into play, by introducing automated measurements, the sample size can be larger, and allows for a more thorough understanding of the quality in that batch.”
Technology is ripe
Van de Zedde says the technology already exists in high-speed sorting systems, but that the team at Wageningen Food & Biobased Research, working under the project title GreenCHAINge, has adapted it to the task of measuring the quality status of produce.
This integrated approach, known as quality phonemics, connects applied research with commercial interests to offer a non-invasive, mostly automated system that not only looks at the outer skin of the product, but also searches for defects within the fruits and vegetables.
The robot is able to test the chemical components, such as the dry matter content and sweetness of the fruit. Everything a human could do, including gently squeezing the fruit to test firmness, but instead of taking just 10 samples, the system is designed to scan and assess hundreds of crates of produce within an hour.
With a more accurate picture of the quality of the produce, better decisions can be made, and hopefully reduce the amount of produce rejected and wasted. This, says van de Zedde, is the real benefit of the technology.
“If we can understand how the quality of produce develops throughout the chain, then you can introduce better scheduling of shipments,” he adds.
“We’re focusing on measuring quality, but aiming for a more efficient logistics chain with less food waste. A serious amount of tropical fruit is being thrown away because of issues around ripeness and ripening.
“This also raises problems for retailers, because a consumer only has to buy say, a bad mango, a couple of times, and they will stop buying that product.”
The team has been working with trading companies such as Total Produce BV, HDG, Univeg, Hillfresh and Fruitmasters to trial the system. Right now, the focus is on the link between a range of sensor data and human, quality-inspection data says van de Zedde, but once that challenge is met, the next one will be to find an engineering firm willing to mass produce the technology.
“Somebody needs to engineer this product and take it to the market, we’re not there yet, but the first measurement systems should be up and running within a year,” he says.
“At The Amsterdam Produce Show and Conference we would like to have discussions with engineering companies and with interested trading companies who want to take this to the point where it could be introduced and sold on the market.
“Already [the system] is getting more efficient, and the measurements are becoming more consistent. It would mean a new way of working, not just from a technology point of view, but throughout the supply chain.
“There is a big need in understanding and exploiting the maximum yield of reliable and robust products starting with growers through to retail [shelves], in which all weak links in the production and transport of food are understood and optimised. Ultimately this will reduce waste, and generate more profits and save costs.”
When, according to the Food and Agriculture Organisation of the United Nations, an estimated 40-50% of root crops, and fruit and vegetables are lost to wastage before they have a chance to make it to the shelves, this is one piece of technology that cannot wait for future application, but needs a place in the market today.
We look forward to seeing you in Amsterdam!