Beeline taught the neural network to control production processes. For example, watch for theft

Beeline’s BigData specialists have developed several solutions in the field of video analytics.

They are aimed at improving the production processes at the enterprise through the use of a neural network trained in unique skills.

There are quite a lot of solutions in the field of video analytics on the market today, however, our models, in addition to the classic functions of entry / exit control, color or face recognition, are able, for example, to detect fire or smoke, control the wearing of personal protective equipment, and also determine the execution of log trimming. The last skill is unique on the market.

The accuracy of the model for determining the wearing of helmets, personal protective equipment and their color has reached 98-99% in both natural and artificial lighting. The model for calculating the number of employees involved was 97% accurate. Also, with maximum accuracy during the pilots, our model determined the length of the remaining log after the trimming operation.

– Director of Big Data Products and Technologies at Beeline Konstantin Romanov

The use of video analytics models will allow you to control product quality, check staff for presence at the workplace, search for strangers, learn about theft attempts, learn about reputational risks and reduce them, as well as increase employee productivity by up to 50%.