How the glassbeam-thingworx integration works

VIVEK SUNDARAM
Friday, December 18, 2015

Machine log data is prevalent in the IoT world, especially for complex machines like medical devices, manufacturing systems, and smart grid architectures; and is a natural outgrowth of building products that require assembly from numerous underlying components. It is very important to collect and mine machine log data to uncover true business insights – similar to mining the proverbial diamond in the rough – and to accrue significant economic value from it.

Mining complex machine log file data opens up an entirely new world of possibilities – with the ability to form correlations, detecting patterns, and performing regressions that were hitherto hard to even imagine. In the IoT world, one doesn’t just have to react to events – one can predict them by detecting patterns in log file data and intelligently react to them based in past treatment of similar events. IoT Analytics platforms adds an entirely new, qualitative edge into the process of converting raw data into meaningful insights, and this evolution naturally lends into next-generation predictive analytics predicated on machine learning.

By leveraging the strengths of Glassbeam and Thingworx it becomes possible to address analytics needs across all data sources, regardless of device complexity. This post touches upon some salient features of the integration.

Real-time data is essential for any operational analytics, where the current state of the system is important to capture. This is often collected through sensors and real time or near real time communication protocols such as SNMP, TCP/IP, XMPP, MQTT etc.

 

A real world example of this is a smart car. The smart car has dozens of sensor driven data sources that are streamed to a central location either periodically or in real time. Thingworx has the ability to model this car as an object with several child objects (chassis, engine, tire, drive train etc.), each with their own properties and data sources. In addition to sensor readings, an advanced smart car would also generate significant events and alarms that occur during continuous operation. Using a Thingworx agent, these logs can then be captured in real-time and pushed to a central repository. In order to process all this data end-to-end in real-time, Thingworx and Glassbeam support Active MQ integration to facilitate real-time streaming of data.

In this approach, the Thingworx capability is enhanced using an FTP extension readily available on the Thingworx marketplace that has been tested certified and maintained by Thingworx (HTTP://MARKETPLACE.THINGWORX.COM/DOWNLOADS/FTP-EXTENSION/). Glassbeam hosts an FTP Server where file drops can be configured. Encryption can be enabled using the FTP Secure protocol.

Thingworx Composer/Visualization

Data transfer is only half the story. The next step is to analyze and visualize the data. Both Glassbeam and Thingworx are capable of visualizing data within Thingworx mashup builder. Glassbeam can be used to generate visualization components as iFrames, which can then be embedded in a Thingworx mash-up, or Glassbeam can act as a REST based web server (Infoserver) for the Thingworx mash-up to query JSON data using HTTPS REST and visualize using Thingworx’s own visualization features.