Trending

All bound points automatically begin trending – no additional setup is required. Once a point is linked to a node and source map, it starts collecting data immediately.

Polling vs Trending

Trending functions independently of polling, which involves several important considerations.

The polling module queries each device for its current point values based on the configured poll_freq. At any given moment, the current value from the node’s perspective is simply the last value that was collected.

The trending module operates on its own schedule, determined by the configured trend_freq (described below). When it’s time to record a trend sample, the trending module uses the last available value from polling.

There’s a tradeoff to consider: more frequent polling leads to more accurate trend data, since values are updated more often. However, the rate at which devices can be polled is constrained by several factors, such as device capabilities, network topology, and the performance of the underlying BMS.

Because polling is a secondary priority to core BMS operations, nodes are designed to respect device and network limits. They will automatically throttle polling frequency if needed to avoid overloading BMS devices or congesting the network.

Trend Frequency

Each point type has a default trending frequency:

• Analog points: every 15 minutes
• Boolean points: use Change of Value (COV) with a minimum interval of 1 minute

Timestamp Alignment

Timestamps are consistently aligned to the top of the hour and the nearest whole minute to simplify data correlation.

For example, an analog point trending at a 15-minute interval will collect at 10:00, 10:15, 10:30, 10:45, 11:00, etc.

Likewise, binary COV data will always align to the nearest whole minute.

Cloud Syncing

Trend data is first collected on edge nodes and then periodically synced to the cloud, where it becomes available for dashboards and APIs.

This approach offers two key benefits:

1. Resilience: If cloud connectivity is lost, edge nodes continue to collect data. Once reconnected, any missed data is backfilled automatically.

2. Efficiency: Trend data is typically very verbose. Edge nodes group and compress this data before syncing, significantly reducing bandwidth usage and storage needs.

Sources synchronize their trend data approximately every hour on a staggered schedule. Generally, trend data should be available within one hour of being collected.