Service for digitizing instrument readings

Computer vision
Find the ability to remotely monitor the operation of equipment and track malfunctions in its work without dismantling and disassembly
Find a solution for reading meter readings and sensors to connect them to the IoT 4.0 system
An industrial company with outdated equipment
Service based on computer vision technology, that "translates" the data of counters, levers, switches into digital form
The ability to "see" the position of the lever, arrow, switch
Reading numbers, symbols on analog, digital displays
To create the service, we used classical and neural network methods of machine learning. Classical methods were used to pre-process and reduce the video format. Then the image was processed by neural network

How it works

The data is structured in the form of reports, graphs
The neural network segments the image, isolates the data
The video stream is transmitted to a remote computer 24/7
We install video cameras in production

Development process

Data collection
Step 01
Image analysis and markup
Step 02
Computer model training
Step 03
Software development
Step 04
Implementation of the computer vision model
Step 05


«Our client needed technical solution that would accelerate the digitalization of production without replacing equipment. At the first stage, we forced his counters and sensors "talk". The second stage is the analysis of the received readings, which will optimize production»
Andrey Ragimov
Project Manager

Development time

2-4 months
depending on the type of equipment and the depth of data processing


Launching neural networks on an edge device with low computing power

Project results

Our client received a ready-made solution, which he was satisfied with. He decided to take further steps to introduce it into production on his own.

Application of the service

Nuclear power station
Drilling rigs
«Our solution can be adapted to almost any production site with outdated equipment. We exclude employees of enterprises from the chain, who now have to check the operation of equipment manually, in uncomfortable and sometimes dangerous conditions.»

Similar projects

The bot that counts pipes on the photo
Neural network for recognizing counter readings
Contact us