Thanks to a growing focus on convenience and customers' increased comfort with automation, demand for unmanned retail solutions has surged. With the sector poised for further growth, Refreshment spoke to Aslak de Silva, CEO of Selfly Store, about the role of artificial intelligence (AI) in the vending industry and how it is shaping the future of the unattended customer experience.

Finnish company Selfly Store, an intelligent vending machine manufacturer and developer, has been focused on AI-driven innovations to empower merchants and enhance customer experiences in recent years. The company was recently acquired by European provider of connected commercial refrigeration Husky Intelligent Fridges.

How has AI changed the vending landscape in recent years?

AI is no longer a futuristic concept for the vending industry – it is the driving force reshaping how businesses operate and how consumers engage with vending solutions.

AI has transitioned vending machines from simple product dispensers to smart retail hubs. AI's ability to analyse data in real time enables dynamic inventory management, personalised user interactions and predictive maintenance. Selfly Store has integrated AI to optimise inventory replenishment, predict customer preferences and reduce operational inefficiencies. These advancements save operators time and improve profitability. Moreover, they have helped more than double the average monthly revenue generated by vending machines and micro-stores.

How does AI enhance the purchase experience for vending machine users?

AI personalises the customer journey by understanding and adapting to user behaviour. Features such as tailored product suggestions, dynamic pricing based on demand and automated expiry management ensure that customers always find fresh and relevant options at the right price.

How can AI tools help operators to drive forward their business?

AI empowers operators by providing actionable insights through tools like Selfly Insights. Operators can analyse sales trends, optimise machine placement and forecast demand with precision. AI also enables predictive maintenance, reducing downtime and ensuring consistent operations. Data collected from historical maintenance issues helps operators to be proactive in fixing problems before they occur. Moreover, year-on-year average revenues per vending machine have increased between 30% and 70%

How do you ensure your clients' data security?

A huge majority of vending machines do not collect any consumer data. For merchant data, at Selfly Story, we use encrypted communication protocols and secure cloud solutions to safeguard customer and operational data. Using a robust, cloud-based platform from known partners ensures we stay ahead of the curve.

Do you use AI for product development and improvement?

Currently, we leverage AI tools, including GPT-based systems, to enhance various aspects of our product development. These tools assist with coding, troubleshooting, debugging and even software design. Additionally, we've developed an internal instance of GPT, fine-tune with our product data using semantic search and the Retrieval-Augmented Generation (RAG) method. This system is particularly valuable for our technical support team, enabling them to provide faster and more accurate assistance.

Looking ahead, AI-powered digital twins have the potential to revolutionise our product development process by stimulating vending machine operations and testing new features before implementation. However, as we currently operate with a single hardware size, the cost-benefit ratio of developing such a model is not yet justified. As our product portfolio expands, this approach will likely play a more prominent role in our innovation strategy.

Does combining the Internet of Things (IoT) and AI have potential within the vending sector?

The combination of IoT and AI is transformational. IoT sensors gather real-time data on machine performance and customer behaviour, while AI analyses this data to drive decision-making. At Selfly Store, this synergy enables predictive stock refills, remote diagnostics and even automated machine learning models that evolve with consumer trends.

What are the potential drawbacks, if any, of AI vending?

While AI offers immense benefits, the quality and accuracy of its output depend heavily on the data provided. With a limited number of transactions per vending machine, there is a higher rish of inaccurate AI conclusions. Operators must constantly monitor, refine and develop AI algorithms. Achieving the best result requires a balance between questioning AI outpurs and trusting its insights.

What is the future of AI and IoT for vending? Are they here to stay, or will traditional, offline vending continue to play a role?

IoT is undoubtedly the future of vending, but traditional offline machines will still serve specific market needs. IoT-enabled vending allows for enhanced connectivity, smarter management and real-time adaptability, making it indispensable in today's fast-paced retail environment.

There will never be a one-size-fits-all situation. Offline vending will have its use cases, but they will be limited. The same applies to AI in general – it is not a solution for every scenario, but its potential is undeniable.

Our conclusion is that vending has a smarter, more agile future with AI and IoT.

At Selfly Store, we are at the forefront of integrating AI and IoT into the vending industry. By empowering merchants with cutting-edge tools, we've created opportunities for growth and transformation. As AI and IoT technologies evolve, the vending industry will become even more adaptive, efficient and user-centric – paving the way for a smarter, more connected future. We're proud to be a part of this revolution, helping businesses thrive in an era where technology meets customer convenience.

A new generation of retail concepts

AI has transformed vending and unattended retail. Once simply a way to sell products in remote locations, now this growing sector offers a streamlined experience for the customer, and an optimised sales and marketing tool for the operator. Below are some recent shelf-service concepts and technologies that are changing the game and changing traditional models...

Cantaloupe Smart Stores address theft and shrinkage

Launched in late 2024, Cantaloupe's Smart Store series is an advanced self-service retail solution designed to "revolutionise the way food and beverage vendors address key challenges" such as labour shortages, theft and shrinkage.

The Smart Stores offer a 24/7 self-service solution with advanced technology and security, designed for retailers, residential buildings, fitness centres, hotel pantries and more. The customer presents payment at the point-of-sale (POS) to unlock the unit, takes the desired item(s), which are added to their cart, then completes the purchase by pressing 'pay' and walking away.

"We've heard from retailers about core customer and operational challenges, which encompass product accessibility, store locations, theft prevention and inventory management," said Jef Dumbrell, chief revenue officer of Cantaloupe. "Our Smart Stores can solve all these issues in traditional retail environments...while extending retail brands into new footprints like airports and college campuses. Additionally, Smart Stores provide a more modern and secure alternative to traditional food and beverage vending machines."

The Smart Store uses weighted-shelves and cameras to monitor inventory and minimise shrinkage, while touchscreens, audio help and visual cues ensure an accessible and seamless experience for the user.

"By addressing core challenges such as theft and labour shortages, our new solutions not only protect our customers' investments, but also open new revenue streams," noted Dumbrell.

Autonomous micro market technology

Micromart, a leader in the convenience retail industry, specialises in autonomous micro market technology aimed at transforming convenience retail.

Created in response to the growing demand for micro markets, Micromart's secured, unattended markets employ AI-powered order processing for fast and secure transactions. Retailers have the flexibility to customise their space with various Micromart units, including the pantry, fridge, freezer and heating tower for hot food capabilities.

The technology features AI-driven order processing and secure payment systems to ensure shopping and prevent theft. It also uses dynamic pricing through electronic price tags and digital video content. Additionally, the Micromart Platform allows operators to manage their business remotely, reducing operational costs and streamlining inventory management.

Yang Yu, CEO of Micromart, said: "After selling over a quarter million items via our KitchenMate food service division in Toronto, we're thrilled to white-label this proven technology for vending, micro market and foodservice operators across the US and Canada".

"Our autonomous micro markets help operators reduce labour costs, minimise theft and optimise revenue, making them an ideal solution for today's economic environment."

Personalised purchases in 90 seconds

In January 2025, vending and automation technology leader Sweet Robo unveiled its Case Bot, an AI-powered vending robot that prints fully customised phone cases in just 90 seconds.

Case BOT allows users to select from a variety of phone case styles, explore an extensive image library or upload their own designs. In under two minutes, this smart robot produces a high-quality, personalised product.

"Case BOT is a revolution in how we think about retail experiences. This isn't just about making phone cases – it's about showing the world what's possible when AI, robotics and creativity come together," said Dennis Branch, chief revenue officer at Sweet Robo.

Aimed at retail, entertainment and entrepreneurial ventures, Sweet Robo's machines are designed to deliver high-impact experiences while requiring less than ten square feet of space.

AI-powered age verification

According to Marcus Tiedt, sales and marketing director at Innovative Technology, the use of AI and biometric technology for self-service payment systems can bring a host of benefits to improve the customer experience.

MyCheckr is Innovative Technology's standalone, automated device for facial age estimation which does not require any integration. The device is perfect for the retail sector selling age-restricted goods or services to help with licensing obligations and Challenge25.

MyCheckr is equipped with a camera and a screen that predicts a customer's age quickly and reliably and can be used in both manned and unmanned settings. The process is customer-centric and seamless, with no specific interaction or registration required.

"AI and biometrics are also increasingly being used to empower personalised in-store shopping promotions to facilitate smart shopping," said Tiedt. "Innovative Technology's age estimation solutions can also be used for digital advertising and consumer analytics and present an idea way to personalise ads and give valuable demographic data to full understand consumer behaviour and provide an enhanced level of customer service."