Background

The customer is a retailer of upholstered furniture. Employees regularly receive extensive lists of products: furniture with a description, reference number and price. They need to update the prices of these products in catalogs and stores across different regions. This needs to be done in the web application and in the mainframe. Previously, about 10 employees made all changes manually: they copied data from spreadsheets and published it on websites.

Business Challenge

Automate the updating procedure and save time of company employees for more important tasks.

Solution highlights

The procedure was automated on the EasyRPA platform. Now employees only prepare the initial spreadsheet and upload it to Google Drive folder. The robot monitors the folder every 20-30 minutes. When it extracts the data, it navigates the web applications and the mainframe and inserts the new prices in the right places. At the end, the robot submits a report to the manager with the results.

Project results

Now the solution is being prepared for putting into commercial operation.

Background

Manufacturers, distributors and retailers sell the same product groups on the same shelves. Market actors need timely and reliable information about the actions of competitors and their own situation against their background. Assessment is performed manually by merchandisers, sales representatives, supervisors and marketers. This process involves a lot of manual labor and is affected by human factor. Marketers do not have time to process all the photos and reports. As a result, decisions are made based on incomplete data.

Business Challenge

The Customer desires to monitor the performance of merchandisers and analyze shelf display information.

Solution

The goods are laid out at the store by store employees or merchandisers. They take photos of product display (realograms) and send them for verification. The IBA Group team obtained several thousand of these photos. Each of them was checked manually, and as a result, almost 100,000 items were marked.

In the course of work, the team faced challenges that needed to be addressed:

• select for analysis only those photos that contain equipment with goods;
• determine the type of equipment correctly;
• subdivide similar products within one category (for example, bottles of water of different sizes);
• process photos of poor quality: light-struck, cropped and with other defects;
• develop a separate process for the introduction of new products;
• come up with a way to merge several photos of long shelves into one using a neural network.

Our experts coped with all the difficulties and developed a SaaS solution to assess photos of realograms:

1. Merchandisers and supervisors upload photos of stock keeping units (SKUs) on the shelves.
2. Photos are uploaded to the IBA Group data center instead of the company’s local server.
3. AI module Plano Checker analyzes the photo, evaluates its quality, searches for SKUs and classifies them according to predetermined scenario. If necessary, several photos are merged into one.
4. Plano Checker module saves reports on the server.
5. Users may access the reports and analytics through a web browser immediately after uploading a photo.

Outcome

Goods Checker shows which goods are in their place on the shelves and which need to be rearranged. It helps merchandisers to work more efficiently and earn bonuses.

The system uses computer vision to determine if the photo shows the equipment containing the product, what kind of equipment it is, and also what types of SKUs are placed on it. Recognition accuracy ranges from 90% to 98% depending on the category of products and the quality of photos.

Automation leaves time to analyze the information obtained. The solution allows tracking surges in demand and actions of competitors 24/7. The system processes data and generates analytical reports for the needs of each manager.

Business challenge

Create a robot that will place orders on the websites of suppliers.

Solution

Customer’s specialists downloaded data from SAP on a daily basis. For example, they prepared lists of automotive spare parts to be ordered from suppliers. After that, the operators visited the websites and placed orders. The number of orders is large; several operators were involved in the process. IBA Group team created a robot that automatically visits the sites of suppliers and places orders.

Result

IBA Group completed a pilot project — created and deployed a robot capable of placing orders instead of an operator.