L’Oreal will recognize plano-grams using neuronets

2018

On January 18, 2018 the Sidisi (CDC) Group and L'Oreal company announced end of a pilot project on automatic recognition of plano-grams using the system of artificial intelligence. Use of neural networks allowed to reduce significantly time of carrying out audit in outlets.

At the beginning of the project training of a system of recognition in the course of which the standard images of products of L'Oréal and pictures of show-windows made by merchandisers in a mobile part of the solution Optimum of ASUMT were compared was provided. A system distinguished goods and plano-grams, using convolution neural networks, compared images to standards and issued information on representation of goods in outlets and compliance of a realogramma with the plano-gram. Results of recognition were used for report generation on different marketing KPI with a possibility of aggregation of data at the top level.

Automatic recognition of the calculation takes only 1-2 minutes that significantly reduces time of filling of merchandising indicators in outlet and by that gives an opportunity to increase the number of visits of employees in day. Medium accuracy of recognition was 93-95% depending on category and type of products L’Oréal.

2016

Alexey Goryachev, the deputy managing director of CDC Group, told at a conference for partners of L'Oreal about solutions for optimization and increase in process performance of routing of the sales representatives and the product delivery developed based on a technology platform the Optimum. In particular, the joint project of L'Oreal with the Ace Target BTL agency on planning of routes of merchandisers using the solutions GIS Optimum allowed to optimize movements of mobile employees by a number of parameters that gave increase in efficiency of working time more than for 10%. Based on results of a pilot project in Moscow, the L'Oreal company plans scaling of the OPTIMUM system for all regions of Russia since May, 2016.