Challenge

Combining the In-Vehicle connectivity to our existing IOT cloud system was a challenge. For Alexa integration we had to create a wrapper around the Alexa intents to map to our IOT format. Also, to make use of the latest Android features, we migrated to the latest Android 11 BSP in Phase 2 of the development. Because of the latest change in the AOSP automotive, the migration was not easy, especially because of the new Dagger architecture, which helps to maintain the code and for migrations to the newer versions of Android.

Solution

The platform of the system is a hypervisor running on a Qualcomm automotive board with Qualcomm BSP backing the Android and Qnx platforms. Qualcomm’s latest 8155P with its Eight core CPU, Adreno 640 GPU and dedicated DSP was enough to cater for the processing power needed for the System. The software stack includes the latest version of AOSP automotive with Custom System UI, Settings, HVAC, Navigation, Media, Connectivity, Cluster and Vehicle services. With custom System UI components using Dagger injection and Runtime Resource Overlays (RRO), we have a more seamless experience of Android Automotive. The System has two physical displays consisting of Cluster and Main display, and a Virtual display on top of the QNX Cluster display. The top Layer of the Cluster is a Virtual display handled by the Android Layer which provides the Weather, Navigation, HVAC and Media views to the Cluster. The more responsive QNX occupies the lower layer of the Cluster display which directly connects to CAN server and provides a smooth and fast graphical experience to the user with instant boot.

Results

The POC was completed successfully with the integration to IOT cloud. With the latest Android AOSP support for customized UI design using Car-UI lib and Qnx Cluster with communication to the CAN network provided a total solution.