RFID Touchless Checkout System
-
Demo
-
Details
This RFID touchless checkout system was the core product of Dynamic Integration Technologies, a startup I co-founded with my close friend, Shakil Jiwa. We started in the summer of 2020 when both of us had our internships canceled due to the COVID-19 pandemic. Shakil was introduced to UHF RFID technology by a colleague, and, despite our limited background in the field, we were intrigued by its ability to uniquely identify tagged items without requiring a direct line of sight and to read up to 1,000 items per second.
Our early tests were successful, and with the growing demand for touchless technology during the pandemic, we designed a touchless checkout and inventory management system using this UHF RFID technology. We built the prototype in Shakil's apartment, tagging items like books, food, and clothing. Our goal was to scan all items in a cart simultaneously as a user approached a checkout screen, eliminating the need to scan each barcode individually. Unlike the Amazon Fresh smart grocery cart, which scans items as users add them to their carts, our system scans the entire cart at checkout. This approach preserves the traditional shopping experience, allowing customers to freely add or remove items from their carts. The only difference is that the items are automatically populated on a checkout screen, eliminating the need for manual barcode scanning.
As we conducted more experiments, we began to see both the potential and the inherent limitations of UHF RFID technology. For example, the accuracy of readings depended heavily on the types of objects in a cart. Carts filled with clothes and books consistently achieved 100% accuracy in our trials, even with over 50 items per cart. However, when items made of metal or containing liquids were present, accuracy dropped significantly—a known limitation of UHF RFID technology. Special tags are available to mitigate this problem, but they are considerably more expensive than standard tags, which cost around $0.07 each. In a store handling hundreds or thousands of items, these costs could substantially hinder the feasibility of the system. Another limitation was the tagging process itself. RFID tags had to be placed on each item, much like a barcode. This tagging had to be done either at the store level or by the supplier, and changing operations at that scale is challenging, especially for stores with large inventories. Despite these limitations, we were committed to the idea of an RFID-based touchless checkout system and believe it held promise. We just needed to find the right niche.
A few months later, we presented our idea and prototype to a VC in Ontario, hoping to secure seed funding and a chance to pilot our project in a retail store. By this stage, we had assembled all the necessary hardware and developed an API to communicate with the reader, making it easy to visualize scanned items. During our pitch, the VC showed considerable interest, though he didn’t seem to fully grasp the system’s limitations and its most suitable use cases. After the meeting, he drove us to a local hardware store and introduced us to the owner. Before we knew it, we had a store willing to test our system, a promise of funding, and the VC’s backing. While this was a great opportunity, it came with some challenges. Our system realistically couldn't function optimally in a hardware store due to the prevalence of metal and liquid materials. Additionally, the sheer volume of items posed a significant hurdle, as tagging all of them would require a significant operational change. We did raise these concerns during the meeting, but sometimes you have to take what you can get and we didn't have any other options at that point.
Over the following months, we established a testbed with approximately 100 items from the hardware store, including challenging items like hardware tools and paint cans. Our checkout system was built using:
- A ThingMagic Sargas 2-Port UHF RFID reader.
- Two strategically positioned RFID antennas to maximize coverage.
- A handheld RFID tag writer.
- A tablet as the checkout screen.
- An android application for writing UPCs to tags.
- An API for initiating and terminating scans with the RFID reader.
- An ultrasound sensor to automatically start scanning as a user approaches the checkout screen.
- A front-end interface for visualizing scanned items and simulating a touchless payment process.
Unfortunately, our progress coincided with a worsening COVID-19 situation, particularly in eastern Canada. The hardware store shut down, forcing us to halt testing. As the lockdowns persisted with no clear end in sight, we began experiencing challenges with our VC. He continued to struggle with the system’s limitations and insisted on making it work specifically for the hardware store, despite our recommendations to shift the pilot to a more suitable environment, such as a clothing or sports store. During this time, we explored other opportunities, but most stores, especially smaller local ones, were either shut down or facing staff shortages. Gradually, without formally deciding to, both of us shifted our focus back to our studies, as we were still completing our bachelor’s degrees.
A few months later, Shakil was accepted into a Ph.D. program at UC Irvine, and I secured a position as a Software Developer at CloudOps. This pretty much marked the end of our venture. Most of the hardware remains in my apartment in Montreal, but it has been years now since we last worked on it. Dynamic Integration, the company we created for this project, has since been dissolved, as both of us have moved on to graduate studies and a new scope of work.
-
Takeaways
On paper, this might be classified as a failed startup, but looking back, I see it as anything but that. Aside from the skills and experience I gained, I had so much fun working on this project with Shakil. From having the idea to building a working prototype, pitching it to a venture capitalist, and testing it in a hardware store with real products—all while balancing our studies—was quite the opportunity, even if we ended up in the negatives financially. I learned that opportunities can come out of nowhere, you just have to be open-minded and willing to try. Anyways, if you've made it this far, thanks for reading.
-
Media