The division of labour is credited for the advancement of the industrial era. It would seem that when it comes to coffee grinding, too much segregation could be stifling progress. Professor Chahan Yeretzian is Head of Analytical and Physical Chemistry at the Zurich University of Applied Sciences. He has dedicated much of his recent work to the science of coffee and the whole value chain, including the study of coffee grinding. In these efforts, he’s noticed an uncomfortable barrier between the engineering work that goes into the technological advancement of coffee grinding equipment, and the empirical taste measurements that will ultimately judge a good cup of coffee. “In all of our work, we’ve focused on making that link between engineering and the results in the cup,” he tells Global Coffee Review. “It’s about understanding what parameters are affecting the cup quality.” Yeretzian says that the link between the surface modification of particles and cup quality has been largely neglected in the academic space. While some in the industry will provide explanations, he says these justifications are largely speculation and incidental, rather than scientifically proven. Over the past few years, Yeretzian has worked closely with grinding machine manufacturers to provide scientific support in the coffee grinding space. One such project has been the use of ceramic materials in the place of steel for grinding plates in disc grinders, and how this may affect cup quality. Because steel is more malleable and susceptible to heat, it has a tendency to shift in the first month of use, and requires more servicing. Ceramic, conversely, will better hold its shape, and thus requires less initial servicing if that shape is accurate. Yeretzian points out, however, that this stability has its downfalls. While a bit cheaper to produce initially, ceramic discs can’t be reused, recycled, or adjusted after they have been cast. Yeretzian plans to release his full findings on the use of ceramic discs later in the year. Throughout his work, he looks at how the different use of materials manifests in the cup. While this project is limited in scope to the hotel, restaurant and café (HORECA) sector, and retail grinders, he points to another area of his work that could have a wider impact in the industrial space. His second focus has been a look at how the temperature of coffee, during the grinding process, affects cup quality. Initial results have shown that the temperature has an end result on the crema. Yeretzian and his team have found that this differentiation in crema is the result of degassing from elevated temperatures during grinding. “What we’re finding is that the temperature has a significant effect on this degassing process,” he says. “The higher the temperature of the grinder, the bigger the degassing effect.” This could have benefits in the industrial space, Yeretzian explains, especially when grinding coffee is destined for capsules. As capsules are immediately sealed, they require some degassing stages. This storage time in degassing silos could be decreased if the grinding process were able to accelerate degassing, thus improving the production line. In his work, Yeretzian says he’s noticed grinding manufacturers are starting to pay more attention to how their equipment fits into the overall production line, and on the final cup of served coffee. “There used to be a lot of talk about the quality of materials used in the equipment, and little talk of cup quality. The link to the coffee was extremely low,” he says. “Since as early as last year, I’ve noticed a greater focus on quality. It’s similar to trends we’ve seen with coffee machine companies, they’ve realised they can’t be removed from the final product.” As for the future of grinding equipment, Yeretzian predicts that the need for precise grinding, especially with coffee capsules, will lead to greater controls in the grinding process. He points to double and triple stage grinders – popular for coffee capsules – as a major advancement in this front. By grinding the coffee in two or three stages, coffee fines (too small particals) can be greatly reduced in the final product. He predicts that the next phase of development could be direct monitoring of size distribution that could be used to automatically adjust the grinder setting. “With the end goal of highest consistency, one direction companies can go will be to analyse the size distribution and readjust the machine to reach a target distribution,” he says. As it turns out, this future isn’t too far off. A study by the Neuronica Laboratory of the Politecnico di Torino and the Research and Development Department of Lavazza has been working on just that. In a paper presented at the 2012 International Joint Conference on Neural Networks, Professor Eros Pasero, Head of the Neuronica Laboratory of the Politecnico of Torino together his team (Luca Mesin, Diego Alberto and Alberto Cabilli) presented their findings on the use of artificial neural networks to control two grinders used at the Lavazza factory. Biological neural networks are the neurons and the synapses inside our brain that allow us to learn and improve functions. Artificial neural networks aim to reproduce that process via computer modelling and programming – essentially a form of artificial intelligence which tries to emulate human reasoning and actions. Data is inputted, then goes through a hidden layer of calculations (the brain), before outputting a result. Just like an experienced machine operator would learn over the years to adjust machine settings to refine the grind, Pasero’s project aimed to use an artificial neural network to replicate that learning and adjustment process. The aim was to have an automated system that would systematically measure coffee particles, and adjust the grinders accordingly. The input data consisted of what operators would normally measure when they suspect some parameter is going out of range. A laser diffraction analyser measured the size and density of the particles. In the study, the authors describe how they were able to generate an optimal level for this input data, to then adjust the distance between the wheels of the two burr grinders when the measured product didn’t meet those optimal limits. The authors are quite positive about their results. “Simulations indicate that the resulting control is in line with the choices than an experienced operator would take,” the study states. “Thus, the results are promising for a future integration of the control system in the production line, to support operators in controlling burr grinders.” Pasero is introducing also the temperature and the humidity among the inputs of the neural network to improve the performance. The authors point to extending their findings to other processes related to the food industry in a national research project called ITACA . And so it would seem that this meeting of the minds between research and industry may prove beneficial for not only the coffee industry, but the advancement of food production as a whole.