Uknow

Artificial Intelligence & Advanced Analytics

Scroll Down

What we do

We thoroughly explore your data and model the available information. If necessary, we also generate additional data acquisition to implement machine learning systems and Computer Vision algorithms.


Our customers will be able to count on reliable and real-time predictions, achieving early detection of anomalies in their process and reducing downtime.

We thus contribute to optimizing your organization’s asset management strategy

SOLUTIONS

Prediction of Anomalies in Infrastructure Networks

Monitoring of Measurement Devices on gas, water networks and power grids.

We develop both software, electronic and telecommunications interfaces to monitor and control measurement devices.

Our clients can thus considerably improve the management of their services by allowing them to include the following functionalities and benefits:

  • Avoid visiting each site to obtain the status of each meter.

  • Visualize and manage information: Detect irregular consumption or network failures by processing consumption factors collected in real time.

  • Evaluation of consumption patterns that enable commercial actions with end users.

  • We implement communications networks based on IoT technology, connecting meters or interfaces in the most convenient way. We also integrate private networks with incumbent telecom operators services
uknow - pic 9
1
3
uknow - pic 2
uknow - pic 1
uknow - pic 04

Fault Identification and Safety Systems (Industry / Transportation)

Through the development and implementation of applications based on Computer Vision algorithms, we offer a wide range of solutions that are customized for the different needs of our customers:

  • Identification of anomalies in real time through image processing to avoid recurrent failures in production lines.

  • Detection of vehicle plates by means of video cameras.

  • Access control of personnel, vehicles and other objects in restricted access areas.

  • Detection of the lack of use of Personal Protective Equipment (PPE).

  • Infringement Detection System on public roads according to customized configuration of spaces such as pedestrian paths, priority at intersections, traffic lights, etc.

Operation optimization in wind power generation and oil wells

Identification and prioritization of workover opportunities in active and inactive wells by analyzing data to increase efficiency in operation and preventive maintenance tasks. We implement predictive models for anomaly detection such as:

  • Automated analysis of concreting problems.

  • Early detection of kerosene wax in wells.

  • Algorithm for detection of water contact with oil (workovers).
6
uknow - pic 08

TECHNOLOGIES

We use Machine Learning and Deep Learning models to develop unique solutions that address our clients’ critical business challenges, achieving significant improvements in their products and services

Deep Learning and Predictions

Image Processing (Computer Vision)

For quality assurance and production process control

Random Forest Model

Trained to identify fault types

Automatic Recognition

Image-based for applications such as automatic inspection and process control.

Recurrent Neural Networks

Supervised Convolutional and Unsupervised Anomaly Detection

NLP

Natural Language Processing

Applied Machine Learning

Predictions with structured data

Linear and non-linear regression models

Classification according to characteristics

Reinforcement Learning

Learning by interacting with the environment

Vehicle detection (YOLO)

Linear and non-linear regression models

Vehicle license plate reading (OCR)

Learning by interacting with the environment