Smart Checkout and Kiosks

Physical retail and food service operations face converging pressures that traditional point-of-sale infrastructure cannot adequately address. According to the National Retail Federation's 2023 National Retail Security Survey, retail shrinkage reached $112.1 billion in losses in fiscal year 2022, up from $93.9 billion the prior year, with the average shrink rate rising to 1.6% of total sales. A Capital One report projected global retail shrink would reach $132 billion by 2024. Self-checkout lanes have compounded the problem, with one industry study finding shrink rates of 3.5% at self-checkout stations compared to just 0.2% for staffed checkout lanes, underscoring the need for AI-enhanced monitoring at unattended terminals.

Labor constraints add further urgency. According to the National Retail Federation, the retail sector had 543,000 unfilled job openings, while the U.S. Chamber of Commerce reported in 2024 that the labor force participation rate remained below pre-pandemic levels despite rising wages. In the restaurant sector, a 2024 National Restaurant Association report found that 47% of operators anticipated increasing reliance on technology and automation to address persistent staffing shortfalls. These workforce gaps directly affect checkout throughput, with studies indicating that 75% of quick-service restaurant customers will leave if seven or more people are in line ahead of them.

The financial stakes extend beyond lost sales. Retailers must balance capital investment in checkout automation against the cost of shrinkage, labor, and customer attrition, all while navigating regulatory requirements such as ADA accessibility standards and PCI DSS payment security compliance.

Smart checkout and kiosk systems employ a layered AI architecture that spans computer vision, sensor fusion, natural language processing, and machine learning to automate or augment the point-of-sale experience. The technology falls into three primary categories: fully autonomous checkout-free stores, AI-enhanced self-checkout kiosks, and smart shopping carts. Each approach uses distinct combinations of ceiling-mounted cameras, shelf-weight sensors, cart-mounted scanners, and deep learning models to identify products and process transactions without traditional barcode scanning.

Computer vision checkout relies on convolutional neural networks trained on millions of product images to recognize items based on appearance, shape, and weight. According to Mashgin, a computer vision checkout provider, the company's systems identify multiple items in under half a second and report accuracy rates exceeding 99.9%. Autonomous store solutions from providers such as Trigo and AiFi use ceiling-mounted cameras combined with shelf sensors to track shoppers and items throughout the store, generating a virtual receipt upon exit. A hybrid approach has gained traction following the April 2024 decision by a major online retailer to replace its fully autonomous Just Walk Out technology in U.S. grocery stores with smart shopping carts, which use computer vision and sensor fusion at the cart level rather than across the entire store ceiling.

For self-service kiosks in quick-service restaurants, AI capabilities include natural language ordering interfaces, personalized upsell recommendations based on order history, and predictive queue management that adjusts staffing in real time. Integration with existing POS, kitchen display, and loyalty systems remains a primary implementation challenge, as does training computer vision models to handle the diversity of products in large-format grocery environments with tens of thousands of stock-keeping units.

Organizations should recognize that fully autonomous checkout remains limited to small-format stores and controlled environments such as stadiums and airports. Large-format grocery deployments have proven technically complex and costly, with one major retailer reportedly spending approximately $1 billion per year in 2019 and 2020 on research, development, and capital expenditures for its autonomous checkout program, according to CNBC reporting in October 2024. The industry is converging on hybrid models that combine AI-assisted scanning with traditional checkout options to balance automation benefits against cost and customer preference.

Mashgin, an AI-powered checkout provider, reported in January 2025 that its computer vision kiosks processed more than 440 million transactions across over 4,000 locations in the United States, Europe, and Australia during 2024. The company's systems were deployed in over 3,000 convenience stores, 145 sports venues, 50 airports, and 50 college campuses, with median transaction times as low as seven seconds. By April 2025, Mashgin announced it had surpassed one billion cumulative transactions, with monthly transaction volume growing 1,233% from three million in March 2022 to 40 million in March 2025. At NFL stadium concessions, fans purchased over 1.4 million items via Mashgin kiosks in 2024, with a median transaction time under 15 seconds.

In the grocery sector, a European discount grocery chain partnered with Grabango to deploy computer vision-based autonomous checkout at a suburban Chicago store in February 2024. The system uses ceiling-mounted cameras and sensors to monitor customer selections, allowing shoppers to exit without scanning items while retaining the option for traditional cashier checkout. In Europe, Trigo Vision reported its technology was live in over 100 stores globally by the third quarter of 2024, serving major grocery chains including a leading U.K. supermarket and large German retailers. In the QSR segment, according to the RBR Data Services Global Self-Ordering Kiosks 2024 report, more than 345,000 self-ordering kiosks were installed worldwide by mid-2023, with a major global burger chain operating more than 130,000 units and reporting a 5% to 6% rise in same-store sales at kiosk-equipped locations.

The smart checkout and kiosk market is segmented into three tiers: autonomous store platform providers, AI-enhanced self-checkout hardware and software vendors, and smart cart developers. According to a 2025 Research and Markets report, the AI-driven retail checkout vision market reached $3.99 billion in 2025 and is projected to grow to $5.05 billion in 2026 at a compound annual growth rate of 26.6%. Grand View Research estimated the broader self-checkout systems market at $5.56 billion in 2025, with North America accounting for 42.8% of global revenue.

Selection criteria for retailers evaluating these solutions include accuracy rates across diverse product catalogs, integration compatibility with existing POS and enterprise resource planning systems, deployment cost per store, ongoing maintenance and software licensing models, and compliance with accessibility and payment security standards. Retailers should also assess whether a vendor's approach requires full-store infrastructure retrofitting or can operate at the cart or kiosk level, as this distinction significantly affects capital expenditure and deployment timelines.

• Mashgin (AI-powered computer vision kiosks for convenience stores, stadiums, and food service)
• Trigo Vision (autonomous store platform using computer vision and shelf sensors for grocery)
• AiFi (camera-based autonomous checkout for micro-markets, stadiums, and convenience formats)
• Grabango (computer vision checkout-free technology for grocery and convenience retailers)
• Zippin (modular autonomous checkout for small-format stores and venues)
• Diebold Nixdorf (self-checkout hardware with AI-powered Smart Vision shrink reduction)
• NCR Voyix (self-checkout systems with computer vision basket recognition and SaaS analytics)
• Toshiba Global Commerce Solutions (hybrid self-checkout kiosks for grocery and retail)
• Instacart Caper Carts (AI-powered smart shopping carts for grocery retailers)