EcoSystem inclusive Computer vision

Engage

Our customer - a leading retail provider in Europe with a shop-in-shop system - had the challenge to monitor the precise sales numbers of their self-service POS.

Monitoring was difficult due to a lack of data provided, as sales were only registered through the different supermarkets where the POS are located.

In a first step, we identified and spoke to the different stakeholders within the company to understand internal processes and how the customer reacts to them. By doing this, we found out, that next to inaccurate sales numbers, inventory and buying behavior were also hard to analyse.

Deliver

In order to address all identified issues and set the company up for long term success, we introduced an advanced monitoring system through computer vision.

With the help of a camera system that is connected to Edge computers, real time data collection and analysis was established.

Our solution provides the customer with a wide range of abilities:

• monitor & analyse buyer behaviour (e.g. which items are often picked up but then put back into the shelf)


• inventory management (e.g. reports about what products are sold at what time)


• inventory replacement (e.g. notification about items low in stock in real time)


• productivity measurements (e.g. at what times a shelf is restocked).

Repeat

We started the project on a smaller scale and introduced the monitoring and analytics setup to 20 POS within Germany.

After fine-tuning the setup and validating the first results, the cameras were quickly installed in an additional 120 POS. We are currently working on rolling out this feature for all european countries our partner operates in.

Next to the setup and installment of the system, we continue to work on optimization and adding additional use cases to the attached analytics platform.

Our goal is to integrate the entire system natively into the existing tech stack of our partner and build advanced analytics functionalities such as pricing and recommendations or automated replentishments based on automatically adjusted thresholds per POS.