Flow is determined to release baristas of a simple time wasting and mundane task with the launch of the Flow grinder, powered by Fiorenzato, that eliminates the worry of adjusting the grinder ever again.
When Scott Nightingale founded Flow telemetry software in 2018, he knew from the start that the grinder would be the “king pin” solution to so many challenges experienced in coffee making.
“Manual grinder adjustment is an absolute nightmare. I spent plenty of time on the coffee machine myself when I owned a café in New Zealand. I’d make lots of tiny little adjustments to the grinder all the time to stay on target because it was just so sensitive – we’re talking about changes to particle size in microns. It’s ever so slight but it matters. It slows down service and you play mental gymnastics trying to average extraction times in your head and figure out what direction and amount to change the grinder,” says Scott, a former software developer and Head of Innovation at Flow.
“If you’ve got any inconsistency in your process. If your machine is not operating right or you’re chasing uncontrolled variables such as thermal expansion in your grinder, you’re not going to consistently achieve your desired recipe output, so let’s start by measuring, then come back and solve the problems with objective data.”
Flow telemetry was designed to connect or “talk” with any commercial coffee machine on the market and understand recipe execution. But the missing piece of the puzzle, Scott says, was connecting Flow to a grinder to control the grind setting and dry dose, two of the biggest contributors to shot time variance. That’s where the new Flow Grinder, powered by Fiorenzato, comes in.
Baristas simply set their preferred shot time (length of shot) and volume target via the Flow interface. The Flow Grinder then talks directly to the Flow Gateway, meaning it can work with any volumetric coffee machine, and monitors the coffee extraction in real-time.
The Flow Grinder uses data analysis from millions of captured espresso extractions in the Flow ‘cloud of data’ to determine if the grind size should be adjusted slightly finer or coarser according to the output of the espresso extraction. It updates the grinder continuously to make micro adjustments according to the set recipe parameters.
Fiorenzato’s F83 E XGi platform and the XGi grind-by-weight system manage the dose and particle size adjustment.
“There’s nothing the barista has to do other than push the button to start the grind. The grinder will adjust if it needs to, and all that’s left to do is keep making coffee,” Scott says.
“Our testing has shown that we can increase recipe execution by 10 per cent in an already high-performing café. The potential for improvement in lower performing sites is even higher. Once we start to see all the shots trending slightly too fast, for example, the grinder can be slightly adjusted to bring the grind back to the ideal target for your espresso recipe.
“We’re not defining the quality of the shot, what we measure, is consistency.”
Brett Bolwell, CEO of Barista Equip, formerly called Barista Technology Australia, is the distributor of Flow. He bought Flow and Fiorenzato together to collaborate and bring this new technology to life.
When Scott saw the grinder that Fiorenzato had developed with self- adjustment capabilities, he knew he could use the smarts of Flow technology to take it to the next level.
“We were really attracted to the way Fiorenzato approached their grinder. I think their grind-by-weight solution is really smart. Managing dose and grind setting is pivotal in the whole coffee production. The only thing missing, was that it needed the data we can provide to be self-adjusting,” Scott says.
“They’ve also built it for speed, which is essential. You can’t afford to slow the barista down while the grinder tares or weights the dose. The Fiorenzato doses based on the previous grind weight, and adjusts when required. This is robust, fast, and more than accurate enough.”
The Flow and Fioranzato development team worked together to create the self- adjusting grinder and four months later, had a prototype working better than they had anticipated.
Scott says the current formula is already outperforming skilled baristas, but he believes it can still be quicker and more accurate. Brands such as Pablo & Rusty’s and Altura in New Zealand are involved in a pilot program to test its functionality and consistency. The data learned from trials will further refine the algorithm to better identify statistical trends and predict changes required even earlier.
“On the first day of testing in a real- world café, the site achieved a higher met recipe score than we’d ever seen from their baristas manually managing the grinder,” Scott says. “This is a sophisticated site. It has some nice equipment with good training and experienced baristas, but the bottom line is that nobody wants to manage a grinder. It’s not very glamorous and it’s not the best part of making coffee.”
For a venue with high output and low skill, Scott says shot to shot variation can be so high that it’s impossible for the barista to know how to manage their grinder to maximise their recipe. Statistics can solve that problem by identifying the setting that will land the most shots in recipe.
Aside from increasing the accuracy of shots, the Flow Grinder will help reduce wasted shots – those thrown away that didn’t meet a high standard, or those served of poor quality that Scott says can “destroy” a customer’s experience and be even more costly.
“We want to free up a barista’s time so they don’t have to worry about the whole aspect of self-adjusting the grinder during service and focus on what they’re doing on the machine instead,” he says.
The Flow Grinder, powered by Fiorenzato, officially debuted at the Specialty Coffee Expo in Portland in April. It will launch to market in Q3 2023.
Barista Equip’s Brett says the exciting thing about the software is its potential.
“Because it’s built an algorithm within it, with millions of coffee shots going through it, it has historical data. And because we have so much data, we are starting to build our own AI within Flow. We’re starting to forward read, predict, and make micro adjustments before it gets to the point where a barista has to throw a shot away,” he says.
“It will use the millions of shots in the database that the program knows about, and the algorithm and AI within the historical data, to forward read and make adjustments accordingly. That’s where we know it’s going be the smartest grind-by-weight auto- adjusting grinder in the world.”
Brett adds that over time, gathered data on the freshness of coffee and its extraction output will help build an algorithm to identify how a shot behaves.
“As coffee ages, the bean density changes and it extracts differently. So, we’ll be able to start building algorithms where we know at two weeks or three weeks old, the output is going to change because the coffee’s older,” he says.
Scott hopes that one day, Flow will be connected to every device in a barista’s operational workflow in the not-so-distant future. The aim is to help improve overall efficiency and speed up service, including flow metres, solenoids, temperature readers, machine mainboards, milk steaming solutions, even automated tampers.
“It is a game changer. I think it’s really, really interesting and disruptive to see what this product can do for the coffee industry,” Scott says.
“There’s a lot of misconception about what are really important factors in making coffee and what are the normal things we’re doing that are undermining the outcome. I just want to bring those very tangible insights to the surface for the barista.
“I’m such a believer in objective feedback and data, and most people don’t really fully understand data yet. It’s not numbers and charts. That’s not at all what we’re about. What we are about is giving you that one little gem of information that’s going to make your game to the next level. My end goal is the same end goal as when I was a coffee drinker, and that’s just to be able to get amazing coffee all the time.”
For more information, visit baristatechnology.com.au
This article appears in the June 2023 edition of BeanScene. Subscribe HERE.