A transformative impact of Artificial Intelligence on diamonds
Diamond grading is constantly evolving and now with Artificial Intelligence (AI) coming into play, it is at the cusp of another big fillip which is also touted to be a major game changer in the near future. Sharing the long and the short of it, Deepa Mishra gives a complete overview of this latest technology.
Earlier diamond grading was a manual process. The professional gemmologists would use their years of experience to grade the stones. Based totally on human perceptions, it was difficult to standardise the diamond grades. There would always be discrepancies between the grades of two gem labs. That’s not all! Cut grading that is essentially assessment of the cut of polished diamonds and has a direct bearing on the diamond’s appearance and value, was sometimes confused with the diamond shape (i.e. round, emerald, princess) which in fact relates to the way the facets of the diamond intersect, creating the stone’s unique dimensions and proportions.
But this changed in 1992, when software that provided automated and computerised measurement of the diamond’s proportions, was first introduced. Even then, nearly 30 years ago, the new computer-based cut grading device gave accuracy that were simply impossible for humans to achieve. Since then, cut grading technology, both – the optical scanners used to create a model of the diamond and the software used to analyse and measure the diamond, had evolved. In the present time, it would not be wrong to say that the levels of precision in cut grading are already at superhuman levels – micron precision, to say the least.
In 2016, once again the precision level went up and the Artificial Intelligence-based colour and clarity grading technologies were announced and two years later, in 2018, the world’s first fully automated, AI driven diamond grading lab was opened. That apart, a few months back, IBM launched its ‘crypto anchor’ project, which uses artificial intelligence and blockchain for identification and authentication of objects. At present, the technology giant is working with the Gemological Institute of America (GIA) to apply crypto anchors in the form of optical imaging data as well as related software (called Verifier) to evaluate and grade diamonds as per the four Cs (cut, colour, clarity and carat weight) and then, capture and store a 3D image of the gemstone.
As of now, the project is still in nascent stage. A system is being built to introduce it into the GIA’s grading process. But if everything goes as planned, soon customers would be able to verify the grade and authenticity of diamonds at the retail level, using a smartphone app. GIA also hopes to integrate this solution to the blockchain by creating and adding records on diamonds and ensure transparency and verifiability of the gems throughout their journey from manufacturer to consumer.
RockHer, a luxury jeweler in Los Angeles, has gone a step further in integrating the AI to their grading process and lately launched an artificial intelligence-enhanced e-commerce site to give consumers a better experience. The brand created a virtual gemmologist called Rosi that uses AI to pulls its data from a wholesale database of currently available GIA-certified diamonds. After the user feeds his details, budget and preferences, Rosi crunches over a million data points and delivers her selections in seconds. Although a human gemologist cannot match Rosi’s breadth and speed of research, it is the human intelligence in the company founders’ decades of experience in the diamond industry that programmed her to consider more than 30 different variables for a diamond, far beyond the typical 4Cs that consumers have come to expect.
However, yet to embrace this new technology, the diamond industry is still figuring out the implications of AI grading on the business. To begin with, here are four ways in which AI can impact the global diamond market.
Always improving its accuracy
AI-driven diamond grading is centred around machine learning. For your information, machine learning involves the data gathering of the device and then, using it to perfect its own algorithms. So by that logic, the more diamonds the AI grading device scans, the more intelligent it becomes and the more accurate its results go on to be. Therefore, with time, AI diamond grading will become more and more accurate and efficient.
Much more than just grading
AI-based diamond grading methodologies can also help manufacturers in sorting their diamond inventory according to market needs. This means, providing a more efficient and consumer-driven supply. For example, diamonds that are graded as I3 for clarity can have many variations. By grading and sorting them using AI grading technology, manufacturers can select a group of I3 diamonds with no black inclusions, in a much faster and more accurate way. Retailers can use the sorting filters to build diamond collections that match the needs of their specific market.
Most of all, boosting consumer confidence
The diamond industry is going through a tough time. Post the busting of the PNB scam, consumers are reluctant to invest in diamonds. The consumer confidence is all time low vis-à-vis diamonds. AI diamond grading can help in restoring customer confidence by providing precise, non-disputable grading results that are backed by unbiased technological processes.
By 2020, up to 85 per cent of customer interaction is expected to be managed by machines and hopefully, with that, the diamond industry restores customers’ faith in the stone once again.