Artificial Intelligence Services

Artificial Intelligence continues to be more actively introduced into various spheres of human life. Many analytical companies (IBM, Deloitte, Gartner) predict the development of AI over the next ten years, as well as the development of new AI-related career tracks. For example, IBM predicts the development of generative AI, which will bring tangible results for businesses. According to McKinsey & Company, applications based on generative artificial intelligence have the potential to contribute between $2.6 trillion and $4.4 trillion annually to the global economy through various business scenarios.

What services do we offer?

Usetech’s many years of experience and expertise in various fields allows us to provide the following services: 

– Computer Vision

– NLP

– Generative AI

– Deep Learning

– Machine Learning

Our cases:

The Resource Consumption Model

Industry: Oil&Gas

Tasks to be solved:

Development of a decision-making system and introduction of best practices as related to resource consumption;

Creation of the Energy Consumption Plan depending on requirements to products quality and output;

Recognition (detection) of oil cracking unit operation modes.

Problem: 

Key project problematics relate to two specific production areas:

— Search for options to increase energy efficiency of production thanks to artificial intelligence technology;

— Implementation of the ability to obtain products with specified quality characteristics due to near-optimal operation mode parameters configuration.

Solution:

As part of the project we conducted data discovery, defined parameters influencing energy consumption and energy resource quality characteristics. 

We developed the existing pattern of energy consumption by the unit with an error of less than 2%, based on the analysis of historical data for 6 months and 3 years.

We developed the predictive optimal consumption pattern, based on historical cases and unit operation modes detection.

Hydrocarbon retrieval with time vector methods

Industry: Oil&Gas

Tasks to be solved:

Hydrocarbon green-fields retrieval;

New wells positioning;

Hydrocarbon reserves assessment;

3D-modeling of hydrocarbon reservoirs.

Problem: 

This is a passive listening technology of hydrocarbon fields retrieval based on low-frequency vibration analysis. It allows the retrieval of green-fields in a sustainable manner, without seismic vibrators and explosions.

Solution:

The solution is based on the liquid and bubble field model: when external production induced or natural fluctuations affect hydrocarbon reservoirs, they switch to their own low-frequency noise generation mode.

Artificial Intelligence technology makes it possible to distinguish spontaneous hydrocarbon fluctuations in microseismic noise of the ground surface at a depth of up to 7 km.

Result:

As a result, a map of parameter abnormalities corresponding to filed limits surface projections is drawn up. Moreover, this method allows the user to divide abnormalities into gas ones and oil ones. 

The project implies migration of measured data to define stacked reservoir accumulation. The result of migration is the 3D-model of hydrocarbon reservoirs to conduct reserves assessment.

The comprehensive approach to ore fields retrieval

Industry: Mining

Tasks to be solved:

Combining the area 3D-map with magnetic survey indications.

Problem: 

Increase in visualization data quality to make decisions on reserves assessment and ore mining.

Solution:

The solution involves combining of fields received as a result of airborne magnetic survey and the 3D-model of the area.

Result:

As a result, 3D topographic maps with an accuracy of up to 0.03-0.05 m and airborne magnetic survey indications are drawn up automatically.

Remote power lines monitoring

Industry: Power

Tasks to be solved:

Development of 3D-models of power lines, wires and environment to assess power lines safety risks.

Problem: 

Key project problematics relate to implementation of mechanisms for remote power lines monitoring and identification of risks of their damage with precise risk area geolocation.

Solution:

The project implementation included the following:

— The receipt of high-definition aerial photographs for the entire line and orthophotographs for the power line section;

— The development of the area digital model across the entire line length with a width of 200 meters, as well as 3D-models of lines and wires;

— Synchronization of power lines 3D-models with space images;

— Identification of hazards from falling trees and their quantity;

— Identification of power line pylons and power lines state monitoring;

— Identification and control of location and state of line wires;

— Detection of visible isolator defects and evaluation of number of isolators at wire attaching point;

— Detection of visible power line pylons defects;

— Check of the earthing wire presence, state and attachment.

Result:

The project results are as follows:

— The system for power line pylons and wires state control and monitoring based on aerial and satellite photographs is developed;

— Functionality to detect and identify defects and risks of power lines damages with exact coordinates fixation is implemented;

— Digital 3D-models of the area are created, and the mechanism of their synchronization with real-time data is implemented;

— The 3D-program was developed to perform comprehensive data processing for line characteristics identification and their automatic inclusion into the data base using the following parameters: 

Forest length within the span;

Distance between outer wires;

Existing width of power line right-of-way;

Main forest area height;

Design power line width from the outer wire;

Area to be extended;

Number of trees;

Span area;

Area to be regularly cleaned from plants.