The world of IoT is getting really excited about the benefits of accessing high fidelity Digital Twins of real world assets. What is a Digital Twin you may ask? In simple terms it is any digital representation of a device or system and how it behaves within its environment. Digital Twins are used through a product’s lifecycle to guide decision making in engineering, manufacture and operations. Chris O’Connor along with some IBM clients, gave a great presentation on Digital Twin recorded during the recent opening of our new IBM Watson IoT headquarters in Munich.
With ‘connected’ IoT devices streaming vast quantities of data to the cloud we are gaining unprecedented knowledge about how systems operate in the real world. To relate this data to our Digital Twin representations and derive the greatest insights, it is important we can relate the data to the higher level context of a manageable asset or system. But how can you do this efficiently when you are faced with a device ecosystem containing tens of thousands of devices of many types, versions, and variants? Well, IBM Watson IoT Platform is solving this problem with the introduction of new beta capabilities we call ‘Device Twins’ and ‘Asset Twins’.
A Device Twin is a logical model of the properties and events coming from a particular sensor. Once defined and instantiated, it provides a consistent means of interacting with a device in a REST-like manner. Because the models can be shared by multiple devices of different makes and models, the IoT application is now insulated from variability and change within the device ecosystem.
An Asset Twin lets you take this concept one step further. Now you can model higher level ‘things’ that group together sensors and devices into a single entity for a developer to interact with. For example your Asset Twin might represent a class of automobile that in turn represents hundreds of underlying sensors monitoring the drivetrain, suspension, & fuel systems. Now let’s say an IoT based fleet management system called on the state & properties of the automobile Asset Twin, it has a simple and consistent way to monitor all its vehicles regardless of make and model.
So in summary, Device Twins and Asset Twins can help you:
- Provide your application developers with consistent interfaces to access event-driven device data in a REST-like manner.
- Normalize data from devices of different makes or models that publish data in different formats.
- Combine event data from several different device types to model any given IoT thing.
So Device Twins and Asset Twins are great news if you are an IoT developer, but this is just the start. This capability is going to keep growing and will become the gateway to exposing real time data to the whole spectrum of Digital Twins across engineering, manufacturing, and operations. Essentially, this will create a Digital Thread that will draw these representations together, and create a powerful integrated set of perspectives driving insight and process improvement across all these disciplines.
For more documentation on how to use Device Twins and Asset Twins in Watson IoT Platform:
And two example scenarios: