The Tesla Energy Platform uses software to give control to customers and utilities when unexpected events happen, such as grid outages, severe weather events, or energy demand peaks. Additionally, it delivers hundreds of thousands of localized greenhouse gas emission and weather forecasts to enable autonomous decision making on the device to meet electric demand reliably, sustainably, and economically.

This talk will explore the architecture of the Tesla Energy Platform including:

• The use of asset hierarchies modeled by distributed sharded actors to track the state of physical devices and their environment.
• The distribution of optimization inputs for batteries to better utilize energy storage at home.
• Functional programming techniques to make the software more composable, reliable, and testable.
• Strategies to decouple critical operational functionality from analytical workflows.
• Trade-offs in edge vs. cloud computing informed by the requirements of a diverse set of smart features.
• Key takeaways on pushing-to-device vs. device-poll models with hundreds of thousands of IoT devices.