The retail and hospitality sectors have been doing a two-steps-forward-one-step-back dance with self-checkout technology. Some store chains, for example, have embraced it only to reportedly discover a related growth in shoplifting, and then reduced availability in response.
Deploying more AI technology at the edge for in point-of-sale (PoS) devices could help these businesses battle the self-checkout shrinkage issues while helping them leverage data-driven insights to create new value-added opportunities.
HP recently announced its AI Accelerator M.2 Card for PoS devices like its HP Engage One Pro G2, Engage Flex Pro G2, Pro-C G2, and Engage Flex Mini G2. According to HP, the card also can be installed in other compatible devices such as commercial PCs and workstations that have an available M.2 slot, providing businesses with the flexibility to leverage AI throughout their environment.
The AI acceleration capabilities at the heart of the M.2 card, which will be available in August, come from Hailo Technologies, which has been targeting its AI accelerator chips at a variety of industries and devices in recent years, including surveillance cameras, single-board computers, and automotive systems.
HP chose the company’s Hailo-10H AI accelerator, which was announced in April 2024, to power the M.2 Card’s ability to generate real-time insights, predictive analytics, and autonomous transactions across retail and hospitality environments.
This move to greater on-device intelligence comes as self-checkout technology is expected to continue growing, with the global market expected to climb from $5.71 billion in 2025 to $18.14 billion by 2034, according to Precedence Research. The theft problem related to self-checkout is part of the broader issue of retail shrinkage, a $100 billion annual problem, according to EY.
“Retail and hospitality are undergoing a profound transformation where the ability to predict, personalize, and optimize in real time is a necessity – and through its innovative AI Accelerator M.2 Card, HP is spearheading that future,” said Andrew Medlin, Senior Director and
Global Head of HP Retail and Industry Solutions, in a statement. “Through our collaboration with Hailo, we’re bringing unprecedented processing power to the front lines, enabling businesses to anticipate customer needs, deliver personalized experiences, and drive next-generation operational efficiency.”
“We’re entering an era where intelligence at the edge is becoming the new standard for retail and hospitality,” added Hailo CEO Orr Danon. “The HP AI Accelerator M.2 Card empowers businesses to reduce shrinkage, tap into the growing opportunity in cashier-less shopping, and unlock powerful capabilities like real-time analytics, enhanced inventory accuracy, and personalized customer engagement.”
Supported applications include:
● Cashierless Checkout – Seamless, autonomous shopping experiences powered by real-time AI processing.
● Personalized Advertisements – Hyper-targeted digital signage that dynamically adjusts based on customer behavior.
● Theft and Shrink Prevention – Real-time monitoring and analytics to detect suspicious behavior and reduce loss.
● Predictive Inventory Management – AI-driven insights that optimize stock levels and prevent out-of-stock scenarios.
A Hailo spokesperson told Fierce Electronics that the company could not disclose the size of the deployment and the number of M.2 cards that will include its technology, but the HP deal is a significant one for Hailo.
“While this isn’t Hailo’s first product to reach the market–our Hailo-8 accelerator is already integrated into hundreds of end products, and the first cameras featuring Hailo-15 vision processors are beginning to roll out–this deployment marks a major milestone,” the spokesperson said. “HP will be the first to launch with our Hailo-10H accelerator, which supports both traditional vision AI models and generative AI. This partnership with a global leader highlights our ability to deliver scalable edge AI solutions and represents a significant step in expanding our presence across the broader consumer technology landscape.”