What if telematics applications could track telemetry from every vehicle in a fleet, and immediately analyze it to identify issues, such as lost or erratic drivers or emerging mechanical problems? What if airlines could continuously track the progress of passengers during their itineraries and proactively respond to delays and cancellations to reduce stress and smooth operations? What if rail operators could detect impending mechanical failures before a derailment occurs?
Applications like these need to simultaneously track the dynamic behaviour of numerous data sources, such as IoT devices and sensors, to identify issues (or opportunities) as quickly as possible, providing operational managers with the best possible situational awareness. The ScaleOut Digital Twin Streaming Service enables the construction of streaming analytics applications to address the challenges. With its new release, this service also now adds the ability to run these applications in simulation both for testing with synthetic workloads and to model complex interactions.
The software “digital twin” model simplifies application development for both streaming analytics and simulations. Digital twins also provide the building blocks needed to separate application design from the orchestration of large-scale deployments with thousands of entities.
Simulate a workload for streaming analytics
To simulate a large population of data sources that send periodic telemetry messages, developers can build a digital twin model for a single physical data source, such as a vehicle in a fleet and then run thousands of digital twins to generate telemetry for all data sources. Acting as a workload generator, they can test a streaming analytics application running in simulation, such as a telematics application, which also can be implemented with digital twins. Once the analytics code has been validated, developers can then deploy it to track a live system.
Many vertical applications can benefit from the simulation of streaming analytics. For example, digital twins can simulate perimeter devices detecting security intrusions in a large infrastructure to help evaluate how well streaming analytics can identify and classify threats. They also can model rail cars in a nationwide rail system to validate streaming analytics that tracks each rail car’s mechanical issues and alert engineers before a derailment occurs.
Simulate a large system with many entities
To aid in operational planning and decision-making, digital twins can also model thousands of entities interacting within a large system. For example, they can implement an airline simulation comprising thousands of airline passengers, aircraft, airport gates, and air traffic sectors. These digital twins maintain state information about the physical entities they represent, run code at each time step in the simulation’s execution, and exchange messages that model interactions. The simulation updates the digital twin state over time to track the results of interactions and provide insights to operational managers.
For example, an airline simulation can measure the impact of flight delays on gate congestion and changes to passenger itineraries. In practice, airlines could use simulations like these to model weather delays and system outages (such as ground stops) and evaluate alternative scheduling decisions that respond to these situations. By running faster than real-time, simulations can help make predictions that assist managers of live systems in their decision-making.
Easily scale simulations
The ScaleOut Digital Twin Streaming Service uses scalable, in-memory computing technology to provide the speed and memory capacity needed to run large simulations with many entities. It stores digital twins in memory and automatically distributes them across a cluster of servers that hosts a simulation. At each time step, each server runs the simulation code for a subset of the digital twins and determines the next time step that the simulation needs to run. The streaming service orchestrates the simulation’s progress on the cluster and advances simulation time at a rate selected by the user.
Building simulation models with digital twins provides a clean separation of application code from the orchestration of the simulation. The streaming service can harness as many servers as it needs to host a large simulation and run it with maximum throughput. It can add new servers while a simulation is running, and it can transparently handle server outages should they occur. Developers need only focus on building digital twin models and deploying them to the streaming service.
Mapping a new path
Digital twins have historically been employed as a tool for modelling the detailed behaviour of a single, complex physical entity, like a jet engine. The ScaleOut Digital Twin Streaming Service takes digital twins in a new direction: simulation of large systems with many interacting entities. ScaleOut Software’s highly scalable, in-memory computing architecture enables it to easily simulate many thousands of entities and their interactions. This provides a powerful new tool for extracting insights about large systems with complex behaviours and gives operational managers important new analytical and predictive capabilities.