Visualizing Customer Behavior with Artificial Intelligence and Actualization of Store Analysis Platform

AI signage can be used for face recognition, visitor count, attribution, and return rates, etc., identifying areas that are often seen by visitors by tracking flow lines, detecting suspicious persons (unusual behavior), detecting shoplifting (not for the purpose of catching the offenders, but to prevent it from happening again by alerting the offenders), and delivering coupons to good customers. Distribution of coupons and special offers to constantly visiting customers

AI Store Analysis Platform

Unmanned store and store analysis platform_image01

○You can stand in front of the signage at the entrance of the store. The AI will attribute and age and gender of the customer and recommend products that may be suitable for that person.

○Head counting allows you to tally how many people have entered or passed by your store or passed by the store.

○When opening a new store at a certain place, the attributes, age groups, and number of passersby on the street can be automatically counted.

Facial recognition with temperature detection

Unmanned store and store analysis platform_image02

Head Count and Flow

・ センサー直下を通り過ぎた人数の測定

Head count measurement + drawing lines
Accurate simultaneous counting of multiple people is possible. By performing entry/exit management, the number of people in the store can be accurately counted, and manual entry restrictions can be replaced by TOF sensors. By using multiple sensors, it is possible to count a wide frontage that cannot be covered by a single unit, and to grasp the direction of movement, walking speed, and whether people have stopped.

Detect suspicious person

It’s got its own algorithms and data in the cloud. It can detect suspicious behavior and
It is possible to prevent shoplifting by “talking” to a target person.

Shoplifting, theft or destruction of cars, destruction of stores, burglary, etc. are common suspicious behaviors. In terms of shoplifting detection, it is important to catch suspicious behavior before it happens and prevent it from happening in the first place.
In conclusion, we can detect people who are likely to shoplift in advance and make them change their behavior. That is the real countermeasure of this tool.

What is suspicious behavior?

【1】Look around restlessly
【2】Wandering around the store with a large bag.
【3】They are mindful of security cameras and other devices.
【4】Be wary of customers who leave without buying any items
【5】Persistent behavior to attract the attention of the store staff

When a shoplifter is spotted with these behaviors, the shopkeeper will be asked to stand close to him or her using body language. This will stop the shoplifting before it starts. You can avoid the unnecessary trouble of calling your parents or the authority.

Detect Shoplifting

The system instantly detects dozens of shoplifting patterns in the server and gives instructions to the staff and manager to prevent shoplifting. The shoplifting detection function is activated when the shoplifting is still carried out.

Unmanned store and store analysis platform_image

Coupon Dispatch

Beacon is a BLE-based technology that allows information to be exchanged between devices that support it.
Specifically, the UUID sent out from a Beacon device can be received by the corresponding smart phone’s built-in APP.

UUID, also known as GUID, is a 128-bit number in the form of 550e8400-e29b-41d4-a716-446655440000. It is built into the originating device and is passed to the receiving device during transmission and reception. By processing this GUID on the web, various information can be displayed on the receiving device. The advantage is that it can display messages specific to a particular location.

Beacon Types

・ Long life stationary beacon (1 year)
・ Coin-type (w/acceleration sensor)
・ Normal Coin Type
・USB type Beacon
・Smallest Type Beacon

Use of this technology is to be able to send out information such as coupons and customers who visit the store.

Unmanned store and store analysis platform_image

Unmanned Store

Unmanned store and store analysis platform

A walk-through, completely cashless store.
The store uses cameras and other information to recognize in real time the customer entering the store and the product in his or her hand, and when the customer stands in the payment area, the product and the purchase price are displayed on the touch panel. When the customer stands at the payment area, the touch panel displays the item and the amount of money to be spent. After the customer has the item, he or she simply checks the display on the touch panel at the exit and makes the payment.