COLIGO EdgeStack

COLIGO EdgeStack is based on real-time Linux and Docker containers. Its plug-in design enables its functionality to be extended while reusing the existing data flow architecture.

Artificial Intelligence

We believe that AIoT can be best put in place in the field, by the Automation Engineer. Indeed, he is the one that best knows the processes, the machines and the automation systems. Therefore, we have developed COLIGO to enable AIoT from the field, with no specific knowledge for data analytics.

ML Models

COLIGO offers out-of-the box ML models that requires nothing else but a simple configuration of the data to analyze.

Anomaly Detection

Anomaly detection identifies abnormal events, malfunctions, or deviations from expected patterns, reducing downtime and enhancing safety and efficiency.

Predictive Analytics

Time series forecasting is the practice of using historical data from sensors and devices to predict future trends, enabling proactive decision-making and enhancing operational efficiency.

Object detection/recognition

Object detection/recognition involves the real-time identification and tracking of objects within video feeds. This enables IoT systems to understand and respond to the visual information, making it valuable for applications like security, surveillance, and smart environments.

Noise detection/recognition

Noise detection/recognition in audio streams allows the real-time analysis and identification of unwanted sounds or disruptions within the ambient audio data. This ensures operational integrity, safety, and quality control by promptly flagging and addressing any undesirable auditory elements.

Anomaly detection/Predictive Analytics

Streaming Learning

Streaming Learning technology is a pivotal approach for industrial IoT applications, as it enables the continuous processing and real-time adaptation of machine learning models as fresh data streams in. This technology is particularly valuable in dynamic industrial environments where sensor data and operational conditions change rapidly, ensuring that predictive and control systems remain agile and effective. Whether it's for quality control or process optimization, Streaming Learning enhances the efficiency and accuracy of industrial IoT applications.

Object & Noise Detection/Recognition

Pre-trained Models

Pre-trained ML models play a crucial role in the efficient recognition of objects and noise within the operational environment. These models, having learned from extensive datasets, provide a valuable foundation for swiftly identifying and classifying objects or anomalies in the industrial setting. By fine-tuning them for noise recognition or object detection, they significantly enhance the precision and responsiveness of industrial IoT systems.

Customer specific ML applications

The architecture of COLIGO EdgeStack was developed for customer specific applications. More complicated ML applications may not be possible with streaming learning and may require specific development, model selection, data identification, model training and deployment. Our engineers are here to support any specific request so please do not hesitate to contact us to share your ideas, needs and requirements.