Artificial Intelligence and Computer Vision Facilitate Restaurant Checkout

Have you ever dined at the self-serve canteen in your company’s headquarters? It might be that, following your break, you have a conference call in which you are scheduled to take part or an important meeting you are required to attend. As a result, you don’t have the luxury of time to leisurely amble through the canteen without a care in the world. Instead, you need to quickly and efficiently select the items you wish to eat, pay for your meal, sit down and eat it, and then race to make your appointment. 

Each of the activities in the canteen require a certain amount of time. Ideally, the bulk of this time should be spent enjoying your meal and conversing with your friends. In many cases, the problem is the speed (or lack thereof) of the checkout queue, because it takes the human operator a significant amount of time to look at all of the items you’ve selected on your tray – in some cases scanning their barcodes – and ring them all up (i.e., enter them into the till) individually. 

<br />A wide variety of food options may be available at a self-serve salad bar (Image source: Benjamin Ashton on Unsplash)

A wide variety of food options may be available at a self-serve salad bar (Image source: Benjamin Ashton on Unsplash)In fact, this activity at a regular checkout can easily take 16 seconds or more (see Proppos: Computer Vision Self-Checkout). While this may not seem like a lot of time, there is a ripple-on-effect as more and more people enter the dining hall. As a result, studies show that a large portion of the time spent in a work canteen—often 15 minutes or more—is devoted to queuing in the checkout line to pay for one’s meal (see Are American workers playing ‘ketchup’ with their lunch breaks?). 

If only there were a better way. Well, it may not be too long before your work’s canteen is equipped with a solution based on the combination of artificial intelligence and computer vision. 

Artificial Intelligence and Computer Vision Save the Day 
Solution provider Proppos has developed an intelligent and autonomous checkout system called Proppos FastPay that integrates a hardware inference platform from ADLINK with an artificial intelligence software solution from Pervasive Technologies. 

<br />Proppos FastPay (Image source: Proppos)

Proppos FastPay (Image source: Proppos)In order to create this system, Proppos was looking for an industrial-grade edge AI platform for several reasons, not least that the operating environment in a restaurant can have high levels of humidity. 

For this particular task, Proppos opted to use ADLINK’s DLAP-201-JT2 industrial-grade edge AI platform, which was designed from the ground up to address size, weight, and power (SWaP) constraints faced by edge AI applications, and which meets all of the requirements laid down by the Proppos team. 

From the software perspective, the development, deployment, and management of the AI checkout system is made easy with ADLINK’s DLAP-201-JT2. The secret is the built-in NVIDIA® Jetson™ TX2, a supercomputer-on-module that not only provides the computing power required for real-time AI inferencing, but also comes with NVIDIA’s comprehensive edge-to-cloud solution for AI applications. 

Should you be lucky enough to have a Proppos FastPay system installed in your canteen, all you will have to do is simply slide your tray under the camera. In just 1.5 seconds, the AI checkout system will automatically identify all of the items on the tray and present the total for your approval. In addition to being 10 times faster than a regular checkout, this dramatically reduces the length of the queue. With a quick tap of your credit card, or by means of a smartphone application, you will be on your way to enjoy your meal with your friends, after which you can return to your post relaxed and refreshed to continue your day. Bon appétit! 

Author: Olivia Chan
Author: Olivia Chan

Senior Product Manager at ADLINK Technology