Reimagining the Wheels – U-Scan Self-Checkout System

self-checkout machine

With the labour shortage in Alberta and other provinces and the desire to cut cost by grocery chains like Safeway and Loblaw, these chains are installing more self-checkout systems. And there is even a Wikipedia entry dedicated to self-checkout and here is the explanation in the benefit section,

“The benefit to the customer is in the reduced checkout time because stores are often able to efficiently run two to six self checkout units where it normally would have had one cashier. Some customers appreciate the ability to not have to deal with anyone, giving an illusion of privacy and anonymity, when in fact the self checkout attendant can track the progress of customers on all machines via a separate terminal known as a RAP (Remote Attendant Post).

The benefit to the retailer in providing self checkout machines is in reduced staffing requirements since one attendant is all that is required to run 4 to 6 checkout lanes at one time.”

Now, let me jump right into the purpose of creating the Reimagining the Wheels series by sharing with you some of my observations that may improve the current U-Scan self-checkout systems.

Cashier at Remote Attendant Post (RAP)

1. Install a webcam overlooking the scale (with sufficient resolution) to provide a photo snapshot or video feed to the RAP (Remote Attendant Post).
2. With a photo of the item (e.g. a vegetable that users couldn’t find a code for) on the scanner/scale, the experienced cashier at the RAP (Remote Attendant Post) can even help enter the correct item code for the customers at the RAP. Thus reducing the trips to the customer checkout post and increasing the efficiency.
3. With photos or video feed of the items, the attendant at the RAP will be able to keep an eye on all checkouts and reducing the number of mis-identified items (pure unintentional mistakes) and cases of frauds (intentional mistakes).
4. I find the pictorial selection menu poorly designed and sometime frustrating to use to pick the correct fruit or vegetables. One way to start solving this problem is by keeping track of the statistics of the precise time customers spent to find the correct “code/picture” for an item. Items that take particularly long implies the menu system is deficient and confusing to the customers and are good candidates to be looked at and improved upon.
5. Note that for every items, we have the data of exactly what the customer bought as it should be the items the customers finally selected (OK reliable) or one that an attendant helped customer picked (more reliable).
6. In conclusion, there is a wealth of information already floating in the system. The trick is to capture them and analyze them. By adding photos capture (with a webcam) and time stamps of the various actions by the users. These data can be extremely valuable in the right hands to fine tune the system to provide better users’ experiences at the Self-checkout systems.

I hope you had some fun reading my first Reimaging the Wheels posting. Feel free to share your feedback, ideas and experiences here. Many thanks.

P.S. During my research, I found this 2004 news item that Optimal Robotics (original maker of U-Scan) sold the U-Scan business to FuJitsu for over US$35 millions.

P.P.S. For verification purpose, the key ideas presented in this posting are time stamped here.