A universal kiosk leveraging visual intelligence
A. Purpose of the invention
Problems with current kiosks. Recently kiosks have been evolving (or being improved) under the slogan of barrier-free design. But when you look at what’s actually been done, usability has actually dropped. The stated goal is accessibility for people with disabilities, but the result is very limited; meanwhile, accessibility for the general public has actually gone backward. Even in terms of payment methods, things are based on credit cards or smartphones, so cash payment is impossible, and the payment range for children and teenagers is extremely narrow. The intent is good, but in fact the on-site usability and practicality have shrunk.
There are roughly two reasons. First, the hardware form, which reuses the existing manufacturing base. And second, the software built on top of it. Because it’s based on the existing payment system, my judgment is that it fails to focus on the fundamental pain points (Pain Point: the inconvenience or difficulty—essentially the pain or problems—that consumers feel about a particular product, service, or in everyday life).
B. Direction of improvement
The ultimate functions of a kiosk for everyone are roughly three:
1. Easy product counting (of items added to the cart)
2. Simple payment
3. Take-out (packaging or delivery)
These three problems are common pain points regardless of disability or non-disability, regardless of age or gender. To solve them, the solutions currently proposed by barrier-free hardware and software are actually limited. It is nearly impossible to reflect all of: the characteristics of every user (with or without disabilities, all ages and genders), the development technology and maintenance of the kiosk providers, and the cost factors and varied spatial characteristics of the small business owners who must purchase and operate them. From the start, the goal itself is too disconnected from the reality of small business owners.
C. Direction of improvement
So I’d like to propose the following.
Through this, the heavy and bulky hardware structure of kiosks could be dramatically downsized.
For example, it could be composed of just a product display stand, a camera, and a structure capable of transmitting and receiving data. By using smartphones for the display required for product scanning, confirmation, and payment, you wouldn’t even need to separately manufacture features such as the height of the payment table or the kiosk’s voice output. That’s because each user’s smartphone is already optimized to their own constraints (the various types of disability). This way, small business owners can minimize the costs they used to spend for occasional users.
1. Easy product counting: Visual intelligence via the camera automatically recognizes products, gathers and organizes their names and prices, and outputs them (visually or by voice). The camera could be installed wherever it can see the products selected by the user—the shelf, the payment table, or the shopping cart. This approach can apply concepts similar to visual models for recycling sorting or fruit-condition classification.
2. Simple payment: Once the user agrees to the output (or finishes any individual edits), a payment link is output. The link could be provided as a QR code or an SMS. The user can pay through their own smartphone. In this process, identity verification and adult-age verification can be handled via the smartphone or SMS (carrier authentication), making it possible to check whether the items being purchased are appropriate for the user’s age. Of course, a separate card reader or cash machine can be used as needed.
3. (Optional) Items recognized (scanned) by the camera move along a rail or similar mechanism into a bag or box and are automatically packaged. For instance, just like putting a book into a library’s book-return machine scans the loan and stores the book in a compartment at the same time, the user puts purchases into a shopping cart or product-scanner (tentative name), and as it scans, the items go into a bag. Once payment is complete, the bagged goods can be picked up. In addition, at the user’s additional request, the system can offer a feature where the packaged items—along with shipping label info—are stored at a separate location and the delivery company is notified to come pick them up.
D. Component summary
1) Product scanning unit
1-1) Product insertion (storage or display) section
1-2) Camera section
1-3) Scan and result delivery section (output or sharing: display, voice, share link, etc.)
2) Payment integration unit
2-1) Payment method selection: cash dispenser, card terminal, payment link (QR code, SMS, SNS, etc.)
2-2) Payment section and payment-result delivery (output or sharing) section
3) (Optional) Packaging/Delivery unit
3-1) Packaging (storing) and pickup of the products
E. References.
Similar cases in other fields
1) Similarity in usage style: library book-return machine
2) Systemic similarity: visual-intelligence-based recycling sorting, or automatic fallen-fruit sorters
F. Linkages
Product registration
1) Traditional method: barcode scanning, entering information such as product name/price/category/expiration date
2) Additional method: parallel execution of per-product data labeling and visual-data preprocessing, automation of manually entered fields (barcode scanning, product name, expiration date, etc.), partial omission of separate input for individual product identification
Sales management, inventory management, ordering system
1) Analysis of product types and sales characteristics by user
2) Inventory management and automatic ordering system based on purchase history
