Welcome to the Titan Control Center Documentation!
This document provides installation and usage descriptions for users, developers and operators.
As part of the Titan research project on Industrial DevOps, the Titan Control Center emerged as a platform for industrial big data analytics. Building upon a scalable and extensible architecture, the Titan Control Center analyzes and visualizes data streams from Internet of Things sensors in industrial production. Supported types of analytics include aggregations, correlation, forecasting, and anomaly detection. The Titan Control Center integrates seamless with the Titan Flow Engine and, thus, supports Industrial DevOps adoption.
The Titan Control Center is implemented as an event-driven microservice architecture. The microservices communicate with each other in a publish-subscribe fashion and most of them employ techniques from big data stream processing. For communication among services, Apache Kafka is used. More details on the architecture can found in our publications:
- A scalable architecture for power consumption monitoring in industrial production environments.
- Goals and Measures for Analyzing Power Consumption Data in Manufacturing Enterprises
Before deploying the Titan Control Center, simply clone this repository or download one of our releases. We provide deployment declarations for Docker Compose and Kubernetes. While Docker Compose is a good fit having a quick look into the software, we highly recommend using Kubernetes for more serious deployments. Instructions for both options are located next to their corresponding declarations:
- Deploying the Titan Control Center with Docker Compose
- Deploying the Titan Control Center with Kubernetes
After deploying the Titan Control Center, the next step you probably want to take is integrating your sensors, such that you analyze or visualize them.
The easiest way to integrate sensor data is using the Titan Flow Engine. It is a low-code platform, which allows domain experts to model industrial data flows in a graphical modeling language. The Titan Flow Engine provides read-to-use bricks (data processing components) for transferring data packets to the Titan Control Center. Likewise, it provides components for reading analytics results from the Control Center and further process them with the Flow Engine.
In addition, data can also be integrated directly via Kafka. In this case, you have to use Confluent Schema Registry along with Kafka and use the Avro schema located in our commons project.
There a basically two ways of interacting with the Titan Control Center:
The Titan Control Center comes with a single-page application visualizing analytics results and allowing for detailed inspection of sensor data. It is structured in four subpages, which can be accessed via the left-hand side panel.
-
The Dashboard page provides an information dashboard, containing several visual components, which show, among others, seasonal (e.g., daily or weekly) pattern, histograms, and the composition of aggregated data.
-
The Sensor Details page is similar to the dashboard, but provides information at sensor-level or for groups of sensors.
-
The Comparison page provides a graphical tool for correlating time series data of multiple data sources. Time series can be arranged in multiple plots with synchronized time domains and allow analyzing certain points in time in detail.
-
The Sensor Management page allows to arrange sensors into groups and groups of groups via drag-and-drop. Note that for accessing sensor from the Dashboard or Sensor Details page, it first to be added to a group.
More details can be found in the Titan Control Center frontend repository. A public show case is provided at http://samoa.se.informatik.uni-kiel.de:8185.
The Titan Control Center can also be used as a platform for custom industrial big data analytics tasks. This can be done by implementing you own microservice, which subscribes to the data streams of other microservices, and deploying it along with the rest of the Titan Control Center.
Imagine, for example, you designed a novel approach for forecasting electrical power consumption. To evaluate it with continuous sensor data, you would only have to package your software in a Docker container and you would automatically be able to also apply it to streams of aggregated data. Also, you would not have to care about data integration or other parts of the system as the Titan Control Center would handle that for you.
Development instructions for individual microservices can be found in their corresponding repositories. For interacting with other microservices and depended systems during development, we suggest using the Docker Compose files.