How we apply Artificial Intelligence (AI) and optimization in vending machines

April 04, 2018 | By Matthew Robards

Take a moment and consider: 90% of the world’s collected digital data was created in the last two years. We are collecting data at an exponential rate, and the opportunities to become more intelligent and more efficient are seemingly endless.

However, we as a population have struggled to adapt fast enough to keep up with is influx of data. In fact, most organisations are currently using only about 12% of the immense amount of data they collect. The remaining 88% of data is currently untapped and is being stored without being put to use. This means that, despite the massive opportunities presented by big data, we are not taking advantage of them.

As a vending machine operator, you probably have more data than you realize: demographics in the area surrounding your vending machine, the purchase history of each machine, the price history, etc. One of the biggest problems facing most vending machine operators or, indeed, consumer packaged goods companies, in general, is not lack of data but rather inability to leverage it to improve the top and bottom line.

Enter HIVERY Enhance: the world’s first artificial intelligence (AI) application for Vending. HIVERY Enhance learns over time by using a company’s own sales data to recommend product assortment, space-to-sales, and price at an individual vending machine level. The result is a set of actionable strategies that allow a vending machine operator to get the best return on each individual machine.

There are three main ways in which HIVERY helps to optimize vending machines.

Harnessing the power of AI

You have more data than you think; and AI uses the data to learn and develop smarter planogram recommendations never possible before. Currently, the issue lies in the ability to interpret this data without the process taking years of your time and energy. This is why the power of AI is so important when it comes to managing and interpreting big data. AI essentially does the thinking for you: it sorts, organizes, and delivers actionable decisions out of the masses of sales data that your company has stored about your products. Because of the speed and efficiency of AI, the process of managing data and building retail strategy becomes realistic and practical instead of overwhelming. And the best part, it learns over time; so it will make only better, smarter and more effective recommendations.

You can read more about this in this great post by Insider-Trends.

Assortment, space and price

HIVERY Enhance learns from the trading history and patterns of every machine in the fleet. It can see which products are selling in each vending machine, using this to construct a unique fingerprint of each machine.

Once machine learning has made sense of your data, Operations Research, or Optimization, is put to use to turn predictions into actions offering the most optimized space for the selected product assortment. It’s all about delivering the best combination of product assortment and space in vending machines at a reasonable investment.

Turning your data into easy-to-follow action steps

At HIVERY, making the output easy for you to understand is as important for us as delivering you the most efficient and optimized information. It is the simultaneous and interdependent recommendations paired with an easy-to-use operation-centric application that sets HIVERY apart. This is achieved with a simplicity of the interface which betrays the underlying algorithmic complexity behind each individualized recommendation. Vending Analytics guides users through a series of what-if scenarios, displaying the key performance indicators customized to each individual operator, for every action the user takes. But to get the maximum improvement, our clients simply click one button – “optimize” – and the recommendations are served up ready for in-market implementation.

Our software converts the findings into easily actionable visible “planograms” and simply feeds into your existing Vending Machine software, whether it is EASITRAXCantaloupe or your own inhouse application. Our solution removes the human factor from the decision-making process and converts it to just “review-action” task.

Vending Analytics is a Software as a Service (SaaS) based solution enabling low IT integration and deployment costs. In fact, according to Computer World, companies that transition to SaaS subscriptions from capital-heavy, on-premise infrastructure installation or maintenance can enjoy an IT spending reduction of more than 15 percent! SaaS solutions mean faster deployments with little to no system upgrade disruption as ongoing upgrades can be installed almost automatically.

A recently published story by talks about how the Coca-Cola Company and related Bottlers have been using Vending Analytics. For Reyes Coca Cola Bottling Co., for instance, it has generated 6 percent of additional revenue and 15 percent of fewer restocking trips. The bottler manages 19,000 vending machines across California and Nevada. You can read more about it on

Subscribe to HIVERY updates