A digital twins PoC
Modeling IoT devices that are interacting with the physical world.
Introduction
This article presents a collaboration between Globant’s IoT Studio and the Data Analytics Studio intended to implement a digital twin proof of concept (PoC) based on building automation sensors as IoT devices. But before digging in within the technical aspects of the implementation, let us begin with the vision and understanding from the IoT Studio of what a digital twin is.
The diagram below presents the mix between the physical and digital world, from which we can define a digital twin as the digital representation within an IoT platform of assets (IoT devices) that enables strategies oriented to impact positively KPI and metrics of operational efficiency through digital solutions.
Now, in the following section we are going to explain how to achieve an implementation of this type, in order to deploy a digital twins based solution integrating an IoT platform, data visualization and augmented reality experiences for building automation sensors.
PoC development components
The following diagram shows the PoC architecture and the development phases covered during the execution, starting from the hardware components, through the integration of cloud based applications used for IoT platform development, data visualization, persistence provider for IoT data and finally augmented reality (AR) experiences development.
Let now us see each component:
1. Hardware integration: A LoRa based building automation sensors integration, in which three types of sensors were connected to a Kerlink Wirnet gateway. The sensors used for this PoC are three references of LoRa End Nodes from Browan: door and window sensors, indoor air quality (IAQ) sensors and presence sensors.
2. IoT platform: The ThingWorx platform is a complete, end-to-end technology platform designed for the industrial Internet of Things (IIoT). It delivers tools and technologies that empower businesses to rapidly develop and deploy applications and augmented reality (AR) experiences. Within ThingWorx, the digital twins modeling was conducted to get in real time and based on events, updates of sensors properties (variables). The following image presents the digital twin representation of an indoor air quality sensor (IAQ).
3. Data persistence: The common point between the IoT platform and the data visualization tool used in this PoC is the persistence provider; in this case a PostgreSQL database in which ThingWorx is storing sensors data and from where the data visualization tool is connecting to get datasets.
4. Data visualization: Amazon QuickSight is a cloud-native, serverless, business intelligence with native ML integrations and usage-based pricing. This was the tool selected to visualize sensors data through dashboards and reports.
Depending on the end user needs, several analysis and charts can be achieved based on the sensor data and the capabilities of the data visualization tool. For example, the chart below shows the behavior of the air quality provided by one sensor per day.
5. Augmented reality (AR): As mentioned before, the selected IoT platform enables the rapid development of AR experiences that integrates IoT data with other resources like images or 3D CAD files to overlay digital content over the real world images captured by a mobile device. The tool used during the PoC for the AR development was Vuforia and in the following screenshot an AR experience triggered by scanning a Thingmark (similar to a QR code) is shown with an image of the sensor, gauges and labels with data taken from the properties of the digital twin.
Conclusion
We started this article with the definition of digital twin and from there, we covered how an implementation can be achieved considering an IoT platform as the starting point to model the assets that are operating and interacting with the physical world. From the IoT platform, the interaction of several applications with assets data is enabled, and range from data storage, to augmented reality experiences, passing through data visualization, to in some cases interact with business apps to add more context to the decision making process.
In the case of this PoC, the idea is to scale up to an IoT building automation use case, in which sensors and building equipment will be modeled as digital twins to enable space analytics, assets management and energy efficiency solutions.
Acknowledgements
I would like to acknowledge Nahuel Carducci — IoT Studio, Christian Castiblanco and Oscar Gonzalez — Data & Analytics Studio, for all the work and support during the execution of this PoC.
