Since Amazon Go's conceptualisation, retail automation has been catapulted into the mainstream, with retailers keeping a close eye on developments in the technology. In this piece, Jenya Beilin, COO, Trigo Vision, off the back of significant funding, tells RetailTechNews how retailers can make the most of this opportunity, and whether the industry is in a position to fully automate their operations in the years to come. 

RetailTechNews: Can you explain what Trigo Vision do?

Jenya Beilin: We have created an advanced computer vision automation platform to provide retailers with an entirely seamless checkout process for their physical stores. Our platform combines a ceiling-based camera network with highly advanced deep-learning technology that tracks the entire store and captures customers’ chosen shopping items during their in-store journey with exceptional levels of accuracy.

We offer retailers options for how to deploy the checkout process. Either their customers pre-register their credit card details, the cameras identify these shoppers at the store entrance, and they can simply walk out once they have finished and their credit cards are charged accordingly. Or, we can provide payment screens at the end of the store that allow customers to confirm their shopping list (without having to scan the products again) and either pay by credit card or cash.

This process drives operational efficiency and reduces costs for brick-and-mortar stores during this transformative and hyper-competitive retail age. Retail automation also allows stores to enhance their customer service and customer engagement levels – customers don’t have to queue, especially in our ‘culture of now’, or get frustrated by dealing with glitches in self-scanning processes. It’s also likely that more staff will be able to better assist and attend to customers on the shop floor.

How does your product differ from other retail automation concepts, such as Amazon Go?

We’d like to clarify that we’re not competing directly with Amazon Go – they are building their own stores with their technology built into the infrastructure. We are giving existing retailers the Amazon Go experience with a solution that offers complete flexibility and scalability for easy and swift deployment into a store of any size, without requiring any type of change to its layout or structure.

We know the way we identify the products, customers, and carts is different, as we use a completely unique data-collection process. We’ve heard that Amazon Go customers have had a few issues with being charged for incorrect items or those they put back on the shelves. Our technology offers exceptional levels of accuracy and this isn’t an issue for us.

We only use basic cameras, whereas Amazon Go uses very expensive cameras and significantly more in number per square footage of the store. Rather than investing in more expensive hardware to enhance the accuracy of image capture, we’ve focused on enhancing the software by leveraging deep learning and AI in creative ways to develop super high-fidelity computer vision tracking.

Amazon Go identifies customers at the entrance with no checkout process at all, so they don’t know what they have been charged for until they check their Amazon account. We offer retailers flexible payment options, including ability for customers to confirm their shopping list before leaving the store, pay by card or cash, without an account or app.

Retail automation has only been adopted on a fairly small scale, to date. What will it take for this sort of technology to be adopted by a major retail store?

Until now, we don’t believe the technology has been advanced, accurate, flexible, or scalable enough for it to be adopted by a major retail store. Any type of automation market has to work with exceptionally high levels of accuracy, particularly in such a direct consumer-facing environment such as retail.

We’re in talks with major grocers globally and the feedback we’ve received has been overwhelmingly positive. We’re confident that Trigo Vision will be the first platform to be adopted by a major retail store.

What is the main challenge for retailers looking to adopt retail automation technology?

The technology needs to be easily deployable into a store of any size, ideally without any changes to existing infrastructure. It also needs to work with exceptional levels of accuracy.

Plus, consumer tastes and shopping habits change according to countries, and so the technology needs to be flexible and accommodate these. For instance, a resident of Silicon Valley will likely be far more comfortable with no checkout process at all, whereas someone in a more conservative state may prefer to confirm their purchases and make payments before walking out with their goods.

Do you think the retail experience will ever become fully automated?

The retail experience will transform dramatically as, over time, much of a store’s operations will be automated. However, the customer engagement and experience will also transform alongside this as retailers will offer a far more customer-oriented and pleasurable experience. It is highly likely the focus of retail manpower will shift to become more customer-centric vs operation-centric, for example interacting with customers will take on a more dominant role than organising products in the store.