An olive crop shortage in Spain and Italy has caused producers of olive oil to become increasingly concerned by the possibility of fraud. Spectroscopy is one possible solution, as Tom Eddershaw discovers
Italy and Spain are two of the biggest producers of olives for olive oil in the world, but disease and drought have put the industry under strain in the past year. This has driven up prices of olive oil to their highest in a decade according to The Independent. The newspaper stated at the start of the year that production would fall from 3.27 million tons produced in 2013/2014 to 2.39 million tons in 2014/2015 with a metric ton costing $4,281.95, a rise of around $600 since the end of 2013.
This price increase has led to a rise in counterfeit olive oil, created using cheap products mixed with chlorophyll and β-carotene in an attempt to pass as expensive goods. Alternatively, fraudsters are diluting the genuine, high quality oils with lower quality substitutes.
This has led producers to work with companies such as Ocean Optics to devise methods to detect these counterfeit goods at various points along the supply chain in order to guarantee that anything that carries the companies brand is authentic.
Henry Langston, product manager at Ocean Optics, said: ‘Anti-counterfeiting or counterfeit detection is not something food companies want consumers to think about. Nobody wants to worry about how safe the food is. But food producers are very conscious of the quality and consistency of their product. Anything that can impact this is on their watch list.
‘There are all sorts of stories, whether it is corruption or graft, if people have the opportunity to sell something as authentic when it’s not authentic, then generally they’ll do it. High value goods are often targeted; expensive delicacies appeal to criminals in the same way as luxury clothes and handbags do to fraudsters in the clothing and apparel industry.’
He explained that within the food and drink industry, there are two areas producers are predominantly concerned about. The first, Langston said, is ‘being able to measure the authenticity of their product in the field: “This thing that is being branded as my product, is it actually my product?”’
Langston said that generally, the most common method is to put an authentication mark on a product’s packaging so that the packaging itself is used to validate the goods. However, if genuine packaging is used to house a counterfeit product, the fake substance could slip through the checks.
To improve the detection rate, Langston said: ‘There are some really interesting applications where people are building molecules that can be identified very specifically and are edible. There are some things starting to emerge within the pharmaceutical and vitamin industries that are FDA-approved and edible [authentication] tags that can be combined on a surface. You then take a handheld analyser, pull a product off the shelf or at any point in the distribution chain, and check if this retailer is selling the authentic product or if there’s some kind of fraud.’
The second area for concern is checking the authenticity of the ingredients coming in from their supply chain which is a high risk area. Langston continued: ‘You can imagine a producer of ready meals buying tremendous amounts of high value ingredients such as olive oil, meats, or cheese. These are things that are particularly prone to being counterfeited,’ and this could potentially lower the quality of the final product, reducing its value.
The issue, as Langston stated, ‘is where the incentive lies for the producer of that ingredient selling it to that big retailer or food manufacturer. There is actually a temptation for them to sell them a lower quality product for the same price [as a higher quality product]. So, the food and beverage producers have a need, or a proven concern, for the detection of fraud in high value products such as olive oil.’
There are a range of techniques used to check samples of liquids; in the past Ocean Optics has looked at NIR transmission and Raman measurements of olive oil, but has also found visible transmission to be very effective. This has led to a range of low cost, simple spectroscopy techniques to identify true olive oil.
The company’s Spark spectral sensor was released earlier this year, and according to Langston its small size and ease of use makes it suited to counterfeit detection. He said: ‘What’s really exciting for us about the Spark is it enables new generations of measurement systems that are smaller and lower cost. Spectroscopy techniques are going to continue to diffuse through different application areas. As the desire to apply spectroscopy becomes more widespread, the need for solutions that can be provided at a cost and size that make them compelling to these broad audience increases.
‘In the past, things like olive oil [counterfeit checking] were done with more expensive laboratory-based equipment. We wanted to see just what differences we could detect with a low cost, basic solution and we were quite surprised at how successful that was. Any time that we can take the system to the sample means that you can not only test more samples more quickly, but you can test it anywhere in the supply chain. You can supply the same test in the supply chain or out in the field. If you want to check something on a supermarket shelf, you don’t want to have to wheel in a giant analyser on the back of a truck.
‘One application we have worked on in the past is for agricultural measurements of potatoes and vegetable crops in Africa where the previous generation before our products had to load bench top analysers onto the back of a truck and drive them across these fields and on horrendous roads. These were things that were designed for the lab so it begged the question: “how often do these things break?”’
Smaller and more rugged designs that use lower power and simpler to use systems are allowing spectrometer companies to further spread their products throughout the food and beverage industry, while the companies using them are able to maintain their brand reputation and ensure their customers are getting what they pay for.
However, away from the luxury goods field, spectroscopy has benefits for analysing liquor. The alcoholic spirit industry is again heavily impacted by fraud; but this time it is a health concern as well as an economic one. When people try to replicate liquor, there is the chance that they will not only generate ethanol but methanol as well, making the drink very dangerous.
Langston explained that there is a similar situation with fraudulent pharmaceuticals; one isomer could be the cure but a very similar one could be fatal. ‘There is a consciousness to not only protect their [the manufacturer’s] reputation and brand, but to protect the safety of their consumer.’
This solution is something that could be used with any kind of liquid sample. So, from one liquid to the next, the user can develop a model around that. Langston said: ‘Flexibility really depends on the sample. We could take what we have done in the spirit market and use the system to authenticate other liquids, such as olive oil. But if we then tried to use this with a solid sample we would have to rethink the optical measurement geometry of the system.’
Ocean Optics’ Spirit Sampler is an example of a portable authentication system developed as a true end user solution. Central to this was development of a complete custom software package that provides users with the ability to develop their own calibration models unique to their brand.
Langston said: ‘To be honest, I believe this is the most challenging part because there are a lot of variables you need to get a range of real life samples and be able to quickly and easily build calibration models. We have worked hard to develop an off-the-shelf solution that authenticates specific liquor brands. This is not only answering “is this whiskey?”, but being able to say “this is a 12-year-old Macallan” and being able to very specifically identify that this is exactly what it says on the bottle.’
Disease was part of the cause for the failed olive crops in Italy according to The Independent; the economic cost for lost crop yields due to pests and disease can be substantial.
Spectrometry can assist with monitoring crops for disease, Chris Van Veen, marketing manager at Headwall Photonics, commented. Farmers are surveying fields using UAV mounted hyperspectral sensors in order to understand why there is growth in some areas but not others, and identify disease or if the field needs more or less water. Hyperspectral images also enable farmers to specifically place pesticides where required, reducing the overall amount used and therefore maintaining soil quality, making parts of the produce more desirable, and reducing the cost of wasted pesticides.
However, there are more applications for hyperspectral imaging including the inspection of poultry carcasses in production facilities. In the past, Headwall has worked with the United States Department of Agriculture to use a hyperspectral sensor to detect disease on poultry and on the surfaces the meat comes into contact with.
Van Veen said: ‘You’re at the behest of governmental agencies who are really putting the clamps down; whether it is the USDA or a global agency, they really want food products to be safer, more wholesome, free of defects, free of diseases, free of anything that is a bad thing.
‘They [governmental agencies] are always pushing the envelope with respect to “how can we look at this stuff better?” They came to Headwall and asked us to help push the science forward and have it adopted in a more wholesale way.’
The USDA is looking at plucked birds that pass the sensor for any kind of defects. The number of carcasses that are allowed to be processed within a facility has recently been increased to 175 birds per minute. This means sensors are required in order to keep the flow rate up and increase the detail of the inspection.
Van Veen continued: ‘These companies want to come up with some other means than RGB imaging that let them see what human eyes might miss. This isn’t to diminish how good those people [factory workers] are, but they [manufacturers] need more quality. Hyperspectral imaging is like a new set of eyes. It’s basically looking at these food products with a high degree of specificity and discrimination.
Van Veen explained that a human inspector is more likely to ask if a product is good or bad, and, if it’s bad, the whole bird is discarded. He said: ‘If a processor can look early on in the process, and salvage some parts of the bird that are not diseased and don’t show any signs of trouble while getting rid of the rest, you maximise your yield.’
Pick of the bunch
Another application reliant on hyperspectral imaging is for grading berries as a harvester unloads a truck from the field. Van Veen said: ‘There are thousands and thousands of pounds on them and the harvester would obviously want top dollar for them. In this case we are talking about cranberries so some of the berries may be very red, some might be less red, some might be white. This is a way for companies to characterise a truck load based on the sample and we see that its composition is X per cent of high-quality, dark red colour, Y per cent of the lesser ones. So with a high degree of specificity, the hyperspectral sensor is grading the crops.’
He added: ‘The last thing you want is for a piece of foreign matter, whether it’s a piece of wood or a twig, to be missed by the human eye.’
As in most of these applications ease of use is important, and the systems work with a predefined software library. However, as in the olive trade, once a new library has been configured, ‘you can take this technology to any kind of high value speciality crop,’ Van Veen said. ‘There’s a lot of money put into planting and harvesting these things, and
for the producer of high value crops, they need to be able to understand how good their crop is.’