Criminal forensics have long used isotope analysis as a powerful geolocation tool to assist their investigations and solve cases. By taking samples of isotope ratios and referencing them against a database of known sources, they are often able to identify the origins of drugs, ballistics, and skeletal remains.
In 2008, Oritain was founded when two chemists from the University of Otago in New Zealand - Dr Helen Darling and Professor Russell Frew - started applying these same principles of analytical chemistry to verify and underpin provenance claims made by food and drink products.
Stew Whitehead, head of client success at Oritain, explained that any natural product - whether plant or animal - will absorb naturally occurring elements and isotopes from its environment - and that these are the key to determining its precise origin.
“An apple, for example, will have been grown in soil with a certain nutrient profile, at a certain altitude and proximity to the sea, and exposed to certain climate conditions. All of these variables cause the trace elements and isotopes to be present in different ratios and concentrations,” he said.
Creating an origin fingerprint
“We measure these ratios and concentrations through mass spectrometry, resulting in a reading that we call an origin fingerprint - essentially an amalgamation of all these elements.”
This fingerprint is the starting point for building a reference database.
“Once we have built up the database for a known origin source, we can take samples of products at various stages in the supply chain to check they match the fingerprint for that product,” he said.
In this way, Oritain can compare any samples to these specifications to establish whether a product is true to its claimed origin, proving that no substitution or adulteration has taken place. It does this by running the readings through algorithms and models it has developed in-house.
Trace elements for greater granularity
Using isotopes in combination with trace elements enables more detailed and accurate origin identification, according to Oritain.
“We have the capabilities to analyse up to 42 trace elements, although not every product will have every element,” said Whitehead.
“Isotopes are good for capturing large origins but you need trace element samples if you want to get down to finer origin classifications,” he explained.
Pilot study on tomatoes
Whitehead cited a pilot study with tomatoes to demonstrate how it is possible to drill down to field or farm level authentication with this combined approach.
“Tomatoes were grown in a hydroponic environment in two, identically designed glasshouses, across the road from one another. The same variety of tomatoes, the same feed inputs and the same hydroponic set-up was used but we were able to differentiate between the two products.”
While this study demonstrated how far science can go, in reality, he said that for most customers, being able to demonstrate evidence of origin at a regional or country level is enough.
The company has numerous examples of how this technology is being applied in the food industry: for underpinning protected geographical indication (PGI) traceability for Welsh lamb and beef; for verifying that baby formula isn’t fake or adulterated; for protecting the reputation of Loch Duart salmon against food fraudsters; for proving Pyramid Valley wine is made from grapes grown in the vineyard it is named after; and for helping Brazilian coffee farmers to prove the provenance of their beans, to name but a few.
“The use cases just keep growing,” noted Whitehead.
EU Due Diligence rules to drive uptake
He predicts that with the imminent EU Due Diligence rules and the growing ESG movement, demonstrating supply chain integrity will only become more important.
"Companies need to be able to prove that there has been no association with forced labour or deforestation. The only way they can be confident of that is if they know where their raw materials came from in the first place,” he said.
In anticipation of this, he said Oritain would be using some of the $57m funding it had recently raised to “scale up” its capabilities for “verticals” such as coffee and cocoa.
Whitehead emphasised that paper-based traceability does not provide enough authentication to satisfy these demands and warned against “over-reliance” on such a system.
“We have seen it time and time again in the industries we work in: supplier declarations saying the origin of the material is X and actually it turns out to be Y. I think there is over-reliance on self-declaration and paper-based traceability - particularly with those long, complex supply chains, more robustness is needed,” he said.
Asked whether Oritain’s technology would be cost prohibitive for verifying the origin of commodities such as those listed in the Due Diligence regulations, Whitehead replied: “We work with organisations of all sizes, from small Manuka honey brands who might only be producing 30-50,000 jars a year, to multinationals who are operating across multiple commodities.”
The affordability of Oritain’s technology stems from the company’s dynamic approach to data - it is constantly developing and updating its reference database to reflect the ever-changing nature of supply chains.
“We are constantly collecting samples from around the world; we have teams in America, Europe and Asia collecting different commodities to add to our reference database. In some instances, we already have databases up and running that our clients can test against. We build those databases at our own expense - that is our IP. After all, the reading is only as good as the data you are comparing it against,” said Whitehead.
He added that Oritain accounts for variability through temporal sampling - in other words, taking samples from the location at different points in time and adding this to its database so that it is constantly obtaining fresh samples.
However, he said that in ten years of doing this Oritain had observed very little variation.
“There might be one or two elements where there is a slight abnormality, but we are not just relying on one element to classify whether a sample is from that origin or not and if there is a significant outlier we can discount that so it doesn’t skew the model.”