"You will receive a notification when the weight changes." Are you looking for such a system?

In the retail sector, such as retail stores and other stores, efficient operation of sales floors and back-of-house warehouses is required because it is necessary to maximize profits while meeting customer expectations by providing the optimal amount of products at the optimal time.

However, in recent years, there have been challenges such as labor shortages due to a decline in the working population, as well as the increased burden of new operations, such as disinfection and social distancing, due to the impact of the COVID-19 pandemic.

To address these issues, Macnica 's Mpression has developed a prototype of an "AI smart shelf" that can efficiently replenish products and items through automatic management. Functions such as visualization of inventory status and notifications of low inventory can be demonstrated in physical stores. By utilizing such AI and IoT technologies, the retail industry can achieve more efficient store operations and customer experiences, and maintain a competitive edge.

Inventory check issues faced by retailers, restaurants, and distribution warehouses

On the sales floors and backyards of retail stores, restaurants, bars, and distribution warehouses, operations are performed to display and stock products on shelves and replenish them as needed in order to deliver products to customers.

Many stores replenish inventory by checking inventory on a regular cycle. This method is also called TBM (Time Based Maintenance). If the check cycle becomes longer, the operational load will be reduced, but the necessary products will not be provided, resulting in opportunity loss. However, if the check cycle is shortened, the operational burden will increase and employees will be forced to bear the burden in the midst of a labor shortage.

Inventory status changes depending on the external environment, such as the influence of weather, surrounding events, and the operation of business partners, so it is extremely difficult to replenish inventory in an optimal cycle. In Japan, which has entered an era of population decline, a shortage of human resources is an issue in all industries. Utilizing digital technology, AI, and IoT technology is effective in maximizing profits with limited personnel and without imposing any burden.

AI smart shelf that converts from regular maintenance to condition maintenance

One approach to solving the issues with TBM is Condition Based Maintenance (CBM). This is a method of diagnosing the condition of equipment and determining the appropriate timing for maintenance. Advances in sensors, cameras, and other IoT devices, as well as computing, have made it possible to achieve CBM without the burden on people.

Macnica has developed a prototype of an "AI smart shelf" to realize CBM for inventory management in retail stores and restaurants. This is a solution that supports the optimization of store operations by using sensors to measure the weight of products and items on the shelf and sending that information to a management app via Bluetooth, enabling the monitoring of quantities and remaining amounts.

This idea was originally conceived from the challenge of replenishing beer servers (barrels) in an izakaya. Since the beer server does not know the remaining amount from the outside, the beer in the server may run out in the middle of pouring beer. Moreover, at that time, the gas filled in the server will blow out, causing inconvenience to customers with noise and scattering of beer, so care must be taken when handling it. We thought that by measuring the weight of the beer server even if we couldn't see it, we would be able to notice the beer just before it ran out.

With the "AI Smart Shelf" management app, if you set the unit weight of the products you handle, you can check how many products are on the shelf. You can also measure and manage the remaining amount of liquid like a beer server. And if you set the alarm reference value for the remaining number and remaining amount, you can receive a replenishment notification e-mail when it falls below that. This eliminates the need for regular patrols and checks, and enables CBM to replenish inventory before it becomes insufficient.

Demo to optimize inventory and remaining quantity management at retail stores and restaurants

Here we will introduce a demonstration of a prototype of the ` `AI Smart Shelf,'' which optimizes the inventory and inventory management of product shelves in retail stores such as supermarkets and restaurants with beer servers. I hope this will give you some hints for actual operation.

System configuration

・ AI smart shelf (weight sensor: large size)
・ AI smart shelf (weight sensor: small size)
・Monitoring app (PC: via Bluetooth)

Assumption

We will conduct a demonstration by imagining a plastic bottle as a tea beverage and a water tank as a beer server (keg).

Demo 1: Monitoring the number of plastic bottles

The first demo is a demo that uses an app to monitor the number of plastic bottles.

Identifying and measuring plastic bottles. You can see that each time the number of plastic bottles on the AI smart shelf increases, the number displayed in the "Tea" column in the app also increases.

Demo 2: Monitoring the remaining amount (weight) of the water tank

The next demo is a demo that uses an app to monitor the remaining amount (weight) of the water tank.

The remaining amount of water is measured and displayed on the app via Bluetooth. You can see that as the water in the water tank on the AI smart shelf decreases, the display in the "Beer Barrel" column in the app also decreases.

Demo 3: Alert email when product inventory falls below a certain number

The last demo is a demonstration of a function that sends an alert email when the product inventory falls below a certain number. This time, we will introduce a case where the number of tea beverages in stock was below the standard value.

The standard value for sending alert emails is set to 3 or less. I'm trying to reduce the number of plastic bottles from 4 to 3.

【result】
The number of displays on the monitoring app has been reduced to three.
Then, I received an alert email on my smartphone informing me of the replenishment of stock!

Possibility of even more intelligent AI smart shelves with advanced AI

With "AI Smart Shelf", we are considering developing smarter store operations by linking with external data. For example, we can determine the age, gender, and member information of visitors from data captured by cameras installed in stores, obtain weather information, temperature, questions, and data on traffic conditions in the surrounding area, and use them to predict demand. It is intended to be utilized for more advanced inventory management.

In addition, you can set replenishment schedules in advance based on weather, temperature, event information, etc., flexibly hold time sales according to inventory status, and analyze communication history with customers, leaflets, advertisements, and sales results. We can also expect contributions to store management other than smart management such as.

Aiming to be a smart store

Macnica 's Mpression utilizes cutting-edge sensors and other devices, as well as AI technology, to create solutions that solve problems in a wide range of industries.

The "AI Smart Shelf" we have developed has the potential to become a solution that can provide better value to customers while making up for staff shortages on sales floors, back yards, restaurants, izakayas, etc. If you are interested in this solution, please contact us.