A new technology from mashgin promises to simplify the cafeteria line. Clients set their food on a scan table and the system identifies all of the food items with an imaging system, looks them up in a product base, calculates the total and charges the client automatically.
From the video demonstration, mashgin’s simplicity should make it a winner with customers. The concept works fast and simply, as it would need to do in this challenging retail environment. There are a number of factors which will influence this concept’s success as a full blown product.
Price – A POS for this type of environment is relatively inexpensive and can probably be had for $1000 – much less if a simple ECR is used. If this device can be had for that price range it could certainly be a winner. If significantly more expensive, it may be tough to win over cost conscious operators.
One could argue that a cashier could be removed from the equation to drive a huge ROI, but it will require a huge leap of faith for an operator to believe that all clients can be trusted to place all items on the scanner. The fear of shrink will probably mean a longer timeline to remove a cashier.
Another potential selling point is utilizing multiple devices with one cashier overseeing them as is done with self-checkout implementations. This is a more viable argument and potentially a better use of the cashier resources.
Payments – My experience with small transactions is that the longest element is tendering, and not scanning. While that seems counter-intuitive, tendering is never completed with a simple universal system in the real world. People pay with cash, credit, debit, and mobile devices now.
It will be important to incorporate payment into the system in a simple way that keeps flow moving. The concept solution shown assumes a mobile solution or use of a credit card with an MSR slot. Apple Pay or NFC cards could work well here. The the MSR card reader slot should be eliminated – that will need to be updated to as EMV is adopted in the US and many international markets. My personal preference is to use an NFC card for food service payments so I can avoid entering a PIN. Expect US fast food organizations to embrace NFC, beacons, or other options more fully once PINs become more common and a simple swipe of a credit card is replaced by people having to enter a PIN at POS; slowing the queues.
Customer Choice – While the system appears entirely intuitive, there’s always a subset of clients that will struggle or reject self service. Some accommodation will need to be made to serve those that don’t wish to use self service. Some customers consider fast service to be good service, while others prefer the human touch. Ultimately the customer is right. Operators will not want to eliminate any potential revenue sources and will want to support any clients that want to eat.
Fraud / Incorrect Reads – The system will require monitoring to avoid shrink. What if two bars of chocolate are set one on top of the other so that the imager sees only 1 item and charges for one? What if the organic coffee is purchased instead of regular? A cup of coffee looks like a circle of black liquid to the imager – it’s impossible to tell without asking the customer or watching them.
Operations – Even though the system works quickly how the flow works in the restaurant environment will require some significant consideration. How many units should be used? Where should they be placed? How many attendants are needed and how are they best deployed? How are exceptions like a system reading a plate incorrectly or an item missing from the image database? How should the queue be arranged for best use of space and simple flow? What if customers have coupons or vouchers or some other discount? How are the units updated? Where does the database reside? Is it simple for local staff to amend and update the product base?
While its certainty not fair to expect a fully developed system from a concept video, it’s important to think through the entire transaction. This concept has much in common with other self service concepts and the issues above are common to all. All of the issues above are certainly addressable with some thought and an operator devoted to working through the solution with mashgin. I would happily use this sort of technology and look forward to seeing a fully developed iteration in a cafeteria line in the future!