Close Ad

The Spoon

Daily news and analysis about the food tech revolution

Robotics, AI & Data

Trax Launches its Computer Vision-Based Inventory System in the U.S.

Singapore-based Trax announced earlier this week that it has officially launched its Retail Watch computer vision-based inventory management system for grocers here in the U.S.

As we’ve covered previously, Trax retrofits stores with customized camera modules on ceilings and on shelves. These cameras snap a picture of the store’s inventory every hour and send those pics to Trax’s cloud, which uses machine learning to analyze each image to see which products are empty and if any products have been misplaced or mis-priced. A report is then sent to store management, who can then dispatch associates to re-stock and replace items.

According to this week’s U.S. announcement], the company has added autonomous robots to its tech stack, but didn’t specify their use or prominence in the overall offering.

A Trax rep told me last year that stores using its system saw a 1 percent bump in sales through improved item stocking. That 1 percent may not sound like much, but it adds up if used across multiple locations for a single retail chain. Plus, 1 percent promises to just come from better workflow of existing practices (inventory management), not adding new ones.

In addition to retailers, Trax’ also counts CPG companies among its customers. For CPG companies, Trax is more of an auditing tool to monitor shelf presence at stores to ensure they are getting what they paid for.

While keeping a close eye on store shelves may not sound sexy, it is becoming an increasingly competitive field. Walmart has developed its own computer vision-based inventory tool. And Bossa Nova, Simbe Robotics and Badger Technologies all make shelf-scanning robots.

For retailers, having more up-to-date inventory data can translate into more revenue. For consumers, it can mean fewer product outages should shocks to the system like pandemic panic buying occur again.

SKS 2020: Grabango Says its Computer Vision Avoids Biases Because it Doesn’t Use Facial Recognition

Grabango CEO Will Glaser said this week that because his company’s technology doesn’t use facial recognition, it can avoid some of the same types of racial biases found in other computer vision and machine learning algorithms. Glaser’s comments came during my fireside chat with him at our Smart Kitchen Summit this week.

Grabango makes cashierless (or as Grabango calls it, “lineless”) checkout technology for grocery retail. Grabango installs hundreds of cameras in the ceiling of a store, and through a combination of computer vision and AI, keeps track of what people put in their basket (or pocket). When shoppers go to check out, they choose a cashier or use an accompanying mobile app that automatically charges for the items kept.

This type of computer vision-based system is also used by other cashierless checkout companies such as Zippin, Standard Cognition and Amazon, and could become more widespread as demand for contactless retail experiences grows.

But the problem with some computer vision + AI systems is that their algorithms can contain the human biases of their creators. As TechCrunch wrote earlier this year, “MIT researchers reported in January 2019 that facial recognition software is less accurate in identifying humans with darker pigmentation.” In a busy retail environment where a lot of people are grabbing a lot of different items all throughout the day, it’s not hard to see how this type of bias could adversely impact people of color.

Glaser recognized some of the problems that can arise when algorithms are only developed or trained on white men in a lab. He made a point of saying that Grabango’s technology does not use facial recognition, and that it has a 99.6 percent revenue accuracy rate. Grabango’s system does follow a person around store, but doesn’t personally identify them (it can, however, stay with a person even after they have put on or taken off a jacket).

Grabango’s system is also being used in real world conditions, so it continues to train its algorithms in a diverse setting. Last month, Grabango announced that it’s integration with Giant Eagle’s GetGo Market in the Pittsburgh area went live. So now we’ll be able to see in a more open environment if Glaser’s claims hold true.

Caper’s New Counter Device Gives Cashierless Checkout to Convenience Stores

Caper, the startup best known for making smart shopping carts for cashierless checkout at supermarkets, today announced its latest product, the Caper Counter.

Meant for convenience and other small store formats, the square-shaped device sits on a counter and shoppers place items in it. Using five built-in cameras along with computer recognition and artificial intelligence, the Counter recognizes all of the items and tallies up the cost. Customers then use a mobile payment system or credit cart to purchase the items and go.

The COVID-19 pandemic has driven demand for more contactless payment options to reduce human-to-human contact in retail. Caper’s pitch to retailers is that its cashierless solution is easier to set up and running than other players on the market. While competitors like Grabango and Zippin require cameras and sensors to be installed around a store, Caper moves all of its cashierless smarts into smaller devices like shopping carts and now the Counter, so there’s no need for an expensive full store retrofit.

Caper’s move comes at a time when the smart shopping cart space is getting crowded. Companies like Veeve, Tracxpoint, and SAI all offer similar functionality to Caper’s cart. And, not for nothing, Amazon recently announced that it has developed its own Dash smart shopping cart and is looking to license out its cashierless checkout tech to other retailers.

When I spoke with Caper Founder and CEO Lindon Gao last week about the Counter and competition in the smart cart space, he was not that concerned. “A couple of main advantages is that we have a hardware and supply chain in China,” Gao said. “We’re already integrated with a lot of these retailers, and being not Amazon certainly helps.”

The Caper Counter is also indicative of the ways technology is making convenience stores more convenient. C-stores are actually a nice proving ground for technologies like cashierless checkout because they are smaller and stock fewer items. They are also stores that people don’t want to spend a lot of time in, so technology that gets people in and out quickly will find a receptive audience.

Gao said that right now the Caper Counter is already in use at several locations with an undisclosed national-level convenience brand partner. Caper wouldn’t reveal specific pricing for its Counter, saying that there is one model where the store pays for the hardware as well as a software fee, with costs dependent on the size of deployment.

Kiwibot will Pick up and Deliver Pizza from Piestro’s Robot Vending Machine

Human-free, contactless delivery is going next-level, thanks to a new partnership announced today between Piestro and Kiwibot.

In a word, this new deal is all about robots. Piestro makes a robot-powered pizza making vending machine , and Kiwibot makes little rover-style food delivery robots. As Piestros come to market, they will feature an integrated mechanism that allows Kiwibots to pluck those pizzas from the machine and cart them directly to your door.

The program isn’t live yet, and there isn’t a specific timetable for when it will happen. Piestro just successfully closed one equity crowdfunding campaign that raised more than $1 million, and is launching a new one with the hopes of raising $5 million more.

But when Piestros do come to market, those living in a Piestro/Kiwibot market will be able to order a pizza from the comfort of your couch through an app, and the Kiwibot will retrieve it and bring it to you. The exact process/app through which you order depends on if the Piestro is one owned and operated by the company, or if it is one that’s branded and licensed by a third-party.

Given the global pandemic, the desire for contactless food delivery is accelerating technologies that power both robotic vending machines and robot delivery services. Not only can robots work around the clock, they also remove a vector of human-to-human transmission. The Piestro/Kiwibot deal takes this concept even further by having a robot interact with the machine so there is no human in the transaction.

The ability to get items from vending machines was a feature that Kiwibot added last year, though being able to hold a boxed 12-inch pizza requires more space than picking up a Snickers bar. I spoke with Massimo de Marco, Piestro CEO by phone this week, who explained that Piestro is working with Kiwibot on a new version of its robot that can carry pizzas.

As noted, there wasn’t any indication of when or where Piestros will be hitting the streets. And even though Piestro and Kiwibot are the first to announce such a daisy chain of robot delivery, you can bet that similar deals will soon follow.

SKS Exclusive: Middleby Unveils the PizzaBot 5000, Which Assembles a Pizza in Under 1 Minute

Lab2Fab, a division of Middleby Corporation, unveiled its new PizzaBot 5000 pizza-assembling machine at the Smart Kitchen Summit today.

The PizzaBot 5000 (or “PB5K” as L2F President, Shawn Lange called it during today’s presentation), is an enclosed robotic system that will spread sauce and cheese on a pizza crust, as well as slice and dispense pepperoni. From there the pizza is removed from the machine, either by a human or by another robot, and placed in an oven for cooking.

The PB5K doesn’t use cartridges for ingredients. Instead, it has big hoppers for cheese and sauce. Pepperoni is loaded as a whole stick and sliced on demand (you can set the thickness). The entire system is refrigerated to keep food safe.

As demonstrated, the PB5K works with three base ingredients (in this case, sauce, cheese and pepperoni), which it can put together in under one minute. The advantages of the PB5K, according to Lange, is that it can crank out pizzas all day, and uses sensors and some computer vision for precise ingredient dispensing, which reduces food waste and save restaurants money.

If this pitch sounds familiar, that’s because that’s also the value prop from Picnic with its pizza-assembling robot. The difference, however, is that Picnic’s system is modular and linear, so it can add as many ingredients by adding more modules. Plus, Picnic has said that its system can be used for foods other than pizza (think: burritos or Subway-style sandwiches).

While not mentioning Picnic directly, Lange did mention that instead of a linear approach to pizza assembly, Middleby has chosen the “clustered” layout for its machine (it’s contained in one cabinet). This gives the PB5K a smaller footprint, with the machine coming it at around 3 – 4 sq. meters.

The PB5K will go into beta in Q1 of next year. Lange said the price was $70,000 for a machine, though the company is exploring a robot-as-as-service model.

UPDATE: An earlier version of this post misstated that the PB5K only worked with pepperoni.

Augmented Reality Bites

This is the web version of our weekly restaurant tech newsletter. Sign up today to get updates on the rapidly changing nature of the food tech industry.

Virtual food hall, meet the augmented-reality restaurant menu. You’ll soon be best friends.

Hear me out.

Over these last few weeks, multiple news bites around virtual food halls have surfaced. These food halls are collections of restaurants that exist online and where meals are only available for delivery and pickup. They are in many ways a natural effect of the pandemic shutting down dining rooms and the restaurant biz going off-premises.

The latest one comes from Lunchbox. This week, the company integrated its online digital order platform into C3’s virtual restaurant brand ecosystem to bring a bunch of different delivery-only eateries under one virtual umbrella.

Being able to order a plant-based burger, chicken, and maybe a rice dish through a single digital interface sounds great until you zero in on that word digital. One of the potential problems with this new wave of virtual food halls is that customers will never have the chance to actually visit these restaurants in person. Your introduction to their food comes in the form of 2D thumbnails you have to scroll through on your phone and squint at to even get an inkling of what you’re about to order. If you’re familiar with the restaurant that’s less of an issue, but most virtual food halls and brands are new, and ordering from them is something of a culinary gamble.

Enter augmented reality (AR), a technology some say is the next great innovation for restaurant menus. Modern Restaurant Management ran a piece this week exploring the possibility of customers using their own smartphones to display 3D models of the food they are about to order. With AR, instead of a small, flat, 2D image, a user could “see” how the dish looks on their table, zoom in on it and view it from multiple angles to get a much better idea of what they’re about to buy.

I should note that the Modern Restaurant Management Post was authored by Mike Cadoux of augmented reality platform QReal. In other words, Cadoux’s has skin in the AR game.

But he makes a good point when it comes to thinking about AR in the context of the new off-premises reality in which restaurants now operate: “Early adoption of AR was hindered by the problem of getting the experience to the customer. People are loath to download apps, and delivery platforms had to service thousands of restaurants, most of which wouldn’t have access to 3D models. Now a restaurant or brand can push their own content to the customer. They would be wise to utilize all the smartphones capabilities and showcase their food with the next-generation of content.”

Spoon Editor Chris Albrecht actually spoke with Cadoux back in August, when QReal released a study with Oxford University’s Saïd Business School that found participants were more likely to order an item if they could view options in AR. “It’s like a test drive for a car,” Cadoux told The Spoon at the time. “Same way when you buy food, you want to think about what it’s like to eat it.”

The tech makes especially good sense for virtual food halls. As I said, these restaurants do not have dining rooms, so customers are relying solely on the digital realm to learn about the food. If, for the sake of argument, Lunchbox and C3 were to integrate AR into their ordering platform, they could better showcase the “fine dining” aspects of their food and in doing so make their meals more appetizing. Everyone else, from Zuul’s virtual-only sandwich chain to Steve Aoki’s pizza brand, could also reap the benefits of AR in the virtual restaurant realm.

AR is not yet mainstream, and its presence in the restaurant industry is still largely forthcoming. But since one pandemic year seems equal to five normal ones, an AR-powered food hall may be closer than we think.

Uber Engineer Says “No” to Uber’s Prop. 22

Californians, take note. One of the things those in the Golden State will vote on come November is Prop. 22, a $180 million ballot measure that would allow third-party delivery services to classify drivers as independent contractors. The measure would effectively override California’s Assembly Bill 5 (AB 5), which was signed into law last year and dictates that Uber, Grubhub, and other gig-economy companies must classify drivers and couriers as employees. 

Classifying them as independent contractors means delivery drivers would lack access to workers comp., paid sick leave, and other benefits W-2 employees receive. It goes without saying that a lot of folks are against Prop. 22. One of them is an employee at Uber.

Kurt Nelson, who’s been a software engineer at Uber since 2018, penned an op-ed at TechCrunch this week that argues drivers should be classified as employees. Nelson, who still makes deliveries for app-based companies in order to understand the gig economy, writes that Uber “refuses to obey” AB5 and instead prefers to “write a new set of rules for themselves” with Prop. 22. 

Among many other notable lines, there was also this gem about the gig economy: “I’ve met drivers who have to sleep in their cars, risk financial ruin over a single doctor’s appointment or go without life-saving medication. There’s no way around it. Uber’s Prop 22 is a multi-million effort to deny these workers their rights.”

You can read the piece in its entirety here. Uber has yet to make any public response to Nelson’s op-ed, so stay tuned.

Restaurant Tech ‘Round the Web

Kitchen United CEO Jim Collins has stepped down to “focus on personal endeavors,” according to Nation’s Restaurant News. Collins played a major role in turning KU into one of the leaders of the ghost kitchen space. Michael Montagano, KU’s former chief financial officer and treasurer, has been named CEO.

Mobile POS platform GoTab launched an integration with hospitality labor management system 7shifts. The combined offering gives restaurant owners/operators the ability to view sales and labor data from the same interface.

Meal prep software company Meallogix announced a partnership with DoorDash this week. A press release sent to The Spoon notes that the deal gives Meallogix’ customers the option of using the third-party delivery service to manage their routes for the last mile of delivery.

Picnic Raises $3M for its Robot-Powered Pizza Assembly

Picnic, which makes a robotic system for assembling foods like pizza, has raised $3 million in funding. Geekwire was first to report the news, writing that “Vulcan Capital, Flying Fish Partners, Creative Ventures, Arnold Venture Group, and others put more money behind Picnic.” This brings the total amount raised by Picnic to $20.7 million.

Picnic’s robot, which was a hit when it debuted at our Smart Kitchen Summit last year, is a modular system that uses computer vision and conveyors to properly dispense ingredients (like pizza toppings). Though it is best known for its ability to assemble pizzas (it can put together 300 in an hour), it has broader applications and could be used to make burritos, subway-style sandwiches and more.

The global pandemic has put a spotlight on food robotics companies like Picnic because they reduce the amount of human-to-human contact when making food, and they also help create more social distance in the kitchen. It also doesn’t hurt that robots don’t get or call-in sick, and they can work around the clock without needing a break. Robots can also work at high volumes. One of Picnic’s first customers was Centerplate Pizza at T-Mobile park in Seattle, though stadiums are pretty much shut down right now.

There are actually a number of companies in the food robot space, angling to bring their automation to restaurant kitchens. Both PAZZI and Piestro are pizza robots. Miso Robotics, just announced the commercial availability of its Flippy ROAR robot. And Blendid’s robot kiosks whip up smoothies.

If you are interested in learning more about Picnic and the future of robots in restaurants, I’ll be hosting a panel with Picnic CEO, Clayton Wood, Dischcraft Robotics CEO, Linda Poulliot, and John Ha, Bear Robotics, next week at our virtual Smart Kitchen Summit. Get your ticket today!

Miso Robotics Announces Commercial Availability of Flippy ROAR

Miso Robotics today announced the commercial availability of its cooking bot, Flippy Robot on a Rail (ROAR). The company also announced that it is working with TimePayment so restaurants interested in adding Flippy to their kitchens can do so with no upfront cost.

Introduced earlier this year, ROAR literally flipped Flippy upside down. Earlier versions of Flippy were installed on the floor of a kitchen, but the redesign inverted that set up, suspending Flippy on rails above ground, thus adding more mobility and freeing up room for human co-workers.

Flippy ROAR has actually already been hard at work in a pilot program at a Chicago-area White Castle since July, and with today’s announcement Miso is looking to expand its footprint to even more QSRs and beyond.

“We really want to drive Miso to serving mom and pops,” Buck Jordan, President and Chairman, Miso Robotics told me by phone this week, adding that the company has received a lot of inbound interest from restaurants of all sizes. “Now it’s easier,” he said.

It’s easier because Miso isn’t charging $30,000 upfront for Flippy any longer. Through TimePayment’s financing options, restaurants of any size can get their own Flippy for $2,000 a month, which includes the setup maintenance and software.

Still, $2,000 is still a lot for an industry that operates on thin margins, especially when those margins are more uncertain than ever. With the COVID pandemic closing many dine-in options, restaurants have had to rely on delivery. But third party delivery services like DoorDash and Uber Eats can charge high commission fees that severely eat into any profits a restaurant might see from delivery.

Jordan said the Flippy can help improve the food delivery experience for restaurants. Flippy’s software now integrates with a restaurant’s ordering platform to help process incoming delivery orders from different third party delivery services. So if a restaurant gets orders at various times from DoorDash, Uber Eats and GrubHub, Flippy’s software can look at when orders come in and pickup ETAs to coordinate cook times for each part of each order. The result is that food doesn’t sit for a long period of time under a heat lamp, waiting for the delivery driver, which should result in fresher food for the end customer.

Flippy could also have the ability to alter the labor economics of a restaurant. Yes, $2,000 isn’t cheap, but a robot can work around the clock, won’t get sick, and has the ability to help create more social distance in the kitchen. If a robot can take over the repetitive and sometimes dangerous tasks of operating the grill and fry stations, restaurants might be able to shift human labor to more higher skilled jobs.

In addition to smart cooking workflows, Flippy is now able to learn how to cook more foods faster. Jordan said that it takes roughly a day of data crunching for it to learn to cook a new item. One such new item Flippy added to its roster is the plant-based Impossible Burgers.

Miso’s announcement comes while the company is in the midst of running its equity crowdfunding campaign. The company has so far raised more than $7.8 million from investors so far.

HWY Haul Brings AI to Automate Fresh Produce Freight and Fight Food Waste

Unless you see one parked outside your grocery store, you probably don’t think too much about the freight trucks that are so vital to getting food onto store shelves. Today, trucking freight, especially produce and fresh foods, from the farm to the store is a pretty complicated and manual process. And HWY Haul wants to change that process with automation and artificial intelligence.

As Syed Aman, cofounder and CEO of HWY Haul explained to me by phone this week, current methods for scheduling shipments require a lot of phone calls and paperwork. Farms need to call brokers to get a trucker, multiple calls must be made to track a load en route, drivers need to stop for temperature checks to make sure produce is kept cold, and once that produce arrives at the store, it can take weeks for a trucker to get paid.

HWY Haul promises to automate this process with what it calls a “managed marketplace.” The company’s cloud-based platform allows farms (or stores or other suppliers) to schedule a vetted driver, determine the cost for each trip, monitor their route in real-time, and keep constant temperature checks (to ensure the food stays cold). Once delivery is made, HWY Haul processes the payment automatically.

In addition to potentially bringing more efficiency to the business of trucking fresh food across the country, Aman said HWY Haul can also help reduce food waste by reducing load rejections. When a load is rejected by the store, that food can wind up in a dumpster. By helping bring loads in on time and always at the correct temperature, Aman said that using HWY Haul can result in fewer of these wasteful rejections.

HWY Haul is actually among a crop of startups working at different points along the supply chain to bring more automation and precision while fighting food waste. AgShift uses computer vision to help establish objective prices for food. Varcode creates blockchain-based thermal stickers to ensure food is kept at the right temperature throughout the cold chain. And Silo automates operations around buying and selling food as well as forecasting supply and demand.

Based in Santa Clara, Calif., HWY Haul has raised an undisclosed Seed round of funding. The company makes its money by charging the shipper a fee per route booked.

Blendid’s New Feature Has the Robot Hold Your Smoothie Until You’re Ready

In addition to making your smoothie, Blendid’s robot will now hold it for you until you are ready to pick it up.

The company posted a video to Linkedin yesterday showing a variety of drinks on its counter. Amidst the colorful assortment of drinks are two small, relatively flat U-shaped brackets. In the video (below) you can see the brackets slide around the surface, pushing drinks into the pickup area.

We’re already seeing an acceleration of interest in food robotics like Blendid’s because of their capabilities and contactless nature. Blendid’s robot can make nine drinks simultaneously, up to 45 drinks per hour and work around the clock. All the ingredients are kept behind glass and sealed away from people, and having a robot slinging drinks means that there is one less human to be a vector for viral transmission.

Blendid is a little ahead of its automated vending bretheren when it comes to contactless retail. In addition to the robot preparing the drinks, ordering is done by mobile phone. Other robot vending services like Chowbotics still rely on touchscreens on the machine itself, which is obviously problematic during a pandemic as dozens of people use the machine in a day.

The addition of Blendid’s hold feature adds another nice bit of social distancing to the company’s offering. Being able to order ahead and have the drink held means customers all don’t need to stand around the machine to order and then wait for their drinks at the same time.

Another interesting bit about Blendid’s new hold feature is how it’s similar to the pucks used by Truebird’s robot coffee barista. Rather than having a robot arms pick up and move drinks around, Blendid’s brackets glide across the counter, sliding drinks into place.

Given that the pandemic is still going strong throughout much of the country, I expect we will continue to see small iterations like Blendid’s across the automated vending space to create even more contactless experiences.

For more on the automated vending machine market, check out my recent report on The Great Vending Reinvention: The Spoon’s Smart Vending Machine Market Report (Spoon Plus Membership required).

Tennant Debuts New, Smaller Brain OS-based Floor-Scrubbing Bot

Tennant, a company that designs and manufacturers cleaning systems, announced today the debut of its new T380AMR Robotic Floor Scrubber, which is powered by Brain Corp.’s Brain OS. The new robot is smaller than other versions of Tennant’s floor scrubbers, allowing the robot to navigate smaller spaces.

The Brain OS is used by a number of different robot manufacturers for a variety purposes including scrubbing and vacuuming floors. Brain-powered robots can autonomously traverse store aisles cleaning, all while avoiding people and other obstacles.

The COVID crisis is placing a spotlight on store sanitation. According to Brain Robots can provide a more thorough cleaning that is also verfiable (you can check the software to see where the robot has been). Additionally, shifting the dull, repetitive work of floor scrubbing over to a robot frees up humans to do other higher-skilled tasks like customer service.

Floor-scrubbing robots are part of a larger move grocery retail is making towards automation. In addition to floor scrubbers, we’re also seeing robots from Bossa Nova and Simbe Robotics autonomously scan shelves to check inventory, and companies like Takeoff and Fabric build out robot-powered automated fulfillment.

T380AMR Ride-On Robotic Scrubber | Product Overview | Tennant Company

Brain-powered floor scrubbing robots are already being used by Walmart, as well as other retailers like Schnuck Markets, Kroger and Giant Eagle.

The news from Tennant today is interesting because a more diminutive robot is built to work in smaller stores with tighter spaces. While no pricing information was made available, presumably this smaller version — meant for smaller stores — would be more affordable and open up autonomy to more stores.

In other words, be ready to see more robots when you go grocery shopping.

OrionStar Launched a New Coffee Robot in China

OrionStar, the robotics arm of Chinese company Cheetah Mobile, jumped into the automated coffee making space last week with the debut of its Robotic Coffee Master.

The Robotic Coffee Master combines two six-axis robotic arms, computer vision, and artificial intelligence to replicate the complex brewing techniques of human baristas such as curves and spirals. Right now the robot makes pour over coffees, using its robotic arms to do things like get the grounds into the filter, set the filter on top of the carafe and pour hot water over the grounds. The Robotic Coffee Master can make a cup of coffee in three minutes.

Unlike other robot coffee makers like Cafe X or Truebird or Briggo, the Robotic Coffee Master is not a full-on enclosed kiosk. Rather it is a squat-looking robot that can be set up in more open environments.

Perhaps once seen as more of a novelty, robotic coffee services could find renewed purpose in a post-pandemic world (whenever that will be). Robots can work around the clock without a break, but also remove a human vector of disease transmission to provide a contactless transaction for customer’s morning coffee.

As noted earlier, the automated coffee space already has a number of players like Cafe X, Truebird, and Briggo here in the U.S., but also international competitors such as Rozum Cafe, MontyCafe, FIBBEE, and Crown Coffee.

The good news is that a lot of people drink a lot of coffee all over the world, so there is plenty of opportunity for everyone involved. In China alone, the coffee market is expected to hit roughly $42.3 billion.

In a phone interview this week, Vincent Li, Head of Global Marketing and Sales, Robotic Solution at Cheetah Mobile, told me that his company is focusing on the Asian market right now. The Robotic Coffee Master sells for roughly $50,000 and the company has already sold 200 units.