This demo simulates a connected vehicle tracking scenario a lot like something we’ve helped one of our customers deploy. In this case, we’ll be tracking car sensor data to show different kinds of maintenance events and faults, and can give individual users the ability to be “tracked” and manually log maintenance faults to see them reflected on the monitoring side of things.
The enterprise end of this equation is a map of London that displays the location and condition of any number of computer- and user-generated cars. The demo simulates the tracking of their location, because the demo is generally delivered to groups of people sitting in a room, which would make for an uninteresting, static and cluttered map.
Car Condition App
The demo also includes a mobile app that people can use to place themselves, in the form of a car marked with their name, on the map among the computer-generated cars. The app gives users the ability to cause and repair four different faults: front flat tire, rear flat tire, engine trouble and busted taillight.
The point is to track the health of each car by sending failures, and fixes, to our backend system, which captures the events and displays them on the Google map. When a user causes a fault their map marker turns red and indicates which of the four faults they generated, along with a “Fix it” button with which they can easily clear the fault.
Behind the Scenes
This is the high-level scaling architecture for the connected car scenario. Solace PubSub+ message brokers runnatively in all clouds; in this case the IaaS/PaaS combination of Pivotal Cloud Foundry running in Google Cloud Platform. The cars connect, they start streaming their sensor data, and the elastic runtime does the work to scale up and down as necessary. All of these components support open APIs and protocols, and they can be auto configured for HA, DR and the kind of reliability you’d expect for enterprise-grade applications.