AI and Machine Learning at the Edge
This article provides an overview of AI and ML at the edge, including implementation, practical applications, challenges, and development tools used to optimize AI models for resource-constrained environments.
Adrian Galeru, our Technical Delivery Manager of the Retail Business Unit, had an opportunity to attend ReTech&Digital on September 14th in Bucharest – one of the leading events in Romania dedicated to Retail Tech and innovations. Experts from the industry gathered together to discuss and showcase the most recent advancements in retail technology.
We are eager to share the most interesting and thought-provoking takeaways Adrian brought back from the event.
Enjoy the reading!
Mobile solutions were unquestionably in the spotlight at the event. These solutions strongly emphasize empowering retail workers by streamlining their activities and increasing overall productivity. Imagine if store staff had access to mobile devices that would make their work more efficient and enable them to reply to consumer inquiries quickly. The main goal is to increase consumer responsiveness and agility in retail operations, resulting in a seamless shopping experience.
Integration of smart labels, an appealing solution to the ongoing problem of price synchronization across online and offline platforms, was one of the key topics highlighted at ReTech&Digital.
Thanks to smart labels, consumers can expect a seamless and open purchasing experience, guaranteeing pricing consistency across different sales channels.
The event highlighted essential developments in AI-based solutions, particularly in macro- and micro-space planning. Planning a store’s overall layout, including the best placement of sections and shelves, is known as macrospace planning.
Microspace planning, on the other hand, focuses on how items are arranged on shelves. Innovative AI-powered planning tools improve client experience while optimizing store layout, ultimately affecting purchase choices.
In the context of retail store infrastructure, offline solutions pertain to the transition towards edge computing, where devices process data locally rather than relying on cloud-based systems. This obviates the need for continuous internet connectivity. Consequently, the associated subscription models are evolving from a volume-based structure, contingent upon the number of images processed, to a fixed-rate subscription regardless of data throughput.
Carrefour introduced the first Romanian retail robot built together with our robotic scaleup Adapta Robotics that span off our internal robotics business unit.
It seamlessly integrates with the store’s operations, extracting accurate data by matching the real-time status of barcodes to the store’s database. Each scan is powered by visual components, including high-definition cameras and depth sensors. With autofocus lenses designed to counteract discrepancies caused by varying shelf distances, the robot collects and assesses pertinent details (prices on both physical and digital labels, barcode positioning, and OSA approximation) directly on the spot.
Processing each high-resolution image using AI algorithms, the retail robot detects discrepancies in price tags, assesses on-shelf availability, and issues alerts for out-of-stock items. This enables the human operator to promptly recognize and address issues, either instantly printing and replacing erroneous labels or replenishing the shelves without delay.
The event made participants more aware of AI-driven sales forecasting, which exceeds the conventional reliance on past sales data. These solutions consider complex market trends in sales and product expiration patterns. Real-time promotions that depended on the time of day or other contextual elements, such as the location and weather, show how data analytics and customer-centric strategies might work together.
Attendees also looked into cutting-edge techniques for attracting customers with personalized promotions. These solutions ensure clients receive tailored offers based on their profiles and interests, enhancing the shopping experience, and promoting brand loyalty. The event also emphasized the value of conversational and generative AI in giving clients accurate and trustworthy answers, reiterating its function in offering excellent service.
Major retail players like Carrefour and Bringo released their bold strategic objectives in line with these developments. By 2026, Carrefour hopes to become a digital retail company, maximizing sales channels to provide clients with seamless solutions, goods, and promotions no matter which channel they use. Bringo, on the other hand, sees itself as a promising concept in online retail: a virtual mall.
Stay tuned as we explore innovative technology developments and business models in further detail, revealing how they have the potential to transform the retail industry and improve customer experiences. We’re thrilled to be at the forefront, seizing the opportunities the retail future promises.
This article provides an overview of AI and ML at the edge, including implementation, practical applications, challenges, and development tools used to optimize AI models for resource-constrained environments.
Exploring why real-time supply chain visibility has become indispensable and how investing in real-time visibility can give businesses a competitive advantage in today’s volatile market landscape.
Sharing trends that will shape automotive software development in the months to come.