meldCX and AOPEN bring machine learning to the checkout

meldCX and AOPEN bring machine learning to the checkou

meldCX & AOPEN to commercialise AI

New York / Melbourne, 14 January 2018 - meldCX, a leading developer of web-based API tools, has successfully demonstrated a machine learning solution that’s ready to be rolled out to supermarket checkouts, speeding up the purchase process and reducing error and fraud.

The solution, based on AI and machine learning at point of sale, has been commissioned for a pilot with New Zealand company Prolife Foods. It’s a self-checkout system that increasingly improves its automatic identification of products, ensuring the correct item is being scanned. The initial trial has involved detecting different kind of nuts. Accuracy improves over time, as the system processes more photographs and learns how to identify the different shapes and sizes.

In addition, customers or solution builders can manipulate their own machine learning models using tools and business rules on the meldCX dashboard without writing a single line of code.

The technology was developed by meldCX using commercial grade devices from AOPEN. It uses meld-intelligence (m-intelligence), an AI system that enables machine learning by way of computer vision and data analysis. This enables it to recognise retail products, learn behaviours, and applies business rules at the edge and in Microsoft’s Azure Cloud.

Dale Tsai, President of AOPEN, says there’s a clear need for businesses to adopt smart vision technology.

"Currently there’s no single place to pull all of this together. We want to democratise AI and machine learning and make these technologies simple and usable for the customer. Inferstem’s mission, by entering this joint venture, is to create a platform for businesses to design, build and deploy powerful solutions, as easily as putting Lego blocks together," Dale Tsai says.

"We see a lot of one-off use cases and concepts. But customers are struggling to actually roll these technologies out at scale. Through AOPEN and meldCX, they’ll be able to commercialise ideas much more easily, and enjoy security, stability, ease of use and development."

In the cloud, m-intelligence trains the solution to recognise products and to apply business rules. Once the machine learning takes place, the rules are pushed down to the devices at the edge.

"meldCX’s hybrid solution combines intelligence from the cloud with processing at the edge, or point of sale. It would take six seconds for an image of the product to go to the cloud and back. By processing at the edge but learning in the cloud, we can detect six products a second, and also apply business rules at the edge. We’ve simplified machine learning over the course of developing the solution," explains Stephen Borg, meldCX CEO and co-founder.

"By applying advanced computer vision algorithms to a camera attached to a scale, the system can see and understand what is going on at the point of transaction and track every item. We can apply this machine learning model across our business and integrate with our supply chain to optimise how we utilise end-to-end information," says Ian Jackson, General Manager, Prolife Foods.

meldCX and AOPEN will showcase the machine learning technology at NRF Big Show 2019 (National Retail Federation) in New York on 14 January, as part of the Intel trade show booth.

Next month, meldCX and AOPEN will also exhibit at ISE (Integrated Systems Europe) in Amsterdam from 5-8 February. The 2019 edition is expected to attract over 1,300 exhibitors and 80,000 attendees, making it the biggest ISE so far.