Mapping Tree Planting – Detecting Repeat Planting Events with the STA Explorer

One of the key features of the STA Explorer is its ability to automatically detect and map tree planting events in the field.

Tree planting is often carried out using planting tools such as shovels, dibble bars, Hamilton planters, and pottiputkis (as outlined in US Forest Service guidance). The STA explorer is designed to attach to these tools. Once mounted, it uses an internal sensor to detect the characteristic movement and impact associated with planting a tree. When that movement is sensed, the device will beep indicating it recorded the event (there is also a physical button on the side of the device, which acts as a failsafe if a planting event isn’t automatically detected).

In ideal conditions, each planting action results in a single, clean event. But as anyone who has spent time planting knows, that’s not always how it plays out in the field.

When one tree becomes multiple events

In tougher environments, a planter may need multiple attempts to successfully install a single tree. The first strike might hit a buried stump, the second might glance off rock or compacted soil, and only the third attempt results in a properly planted tree.

From a data perspective, each of these attempts is recorded. The STA Explorer captures every trigger event, and this is reflected in the GPS_trigger field:

• 1 = accessory trigger (e.g. sprayer)
• 2 = button press
• 4 = planting sensor

So in the example above, all three planting attempts would be logged as valid trigger events.

How the system distinguishes “real” plantings

To make the data usable, the platform applies an additional layer of logic through the Mapping field. Every event is still recorded, but it is classified as either:

• Mapping = 1 → accepted planting event
• Mapping = 2 → filtered (repeat/failed) attempt

Only events where Mapping = 1 are displayed in the web map by default. The rest remain in the dataset behind the scenes, available for more detailed analysis if required.

Distance thresholds and why they matter

This filtering is primarily controlled by a distance threshold in the software.

By default, this is set to a 1.2 metre radius. If multiple planting events occur within this distance of each other, they are assumed to be repeat attempts at the same planting location. The system keeps one as the accepted event and flags the others as failed attempts.

This works well in many scenarios, but planting density plays a big role in how effective this approach is. In high-density planting situations, trees are placed close together. A large threshold risks incorrectly filtering out legitimate plantings. In these cases, reducing the threshold, or removing it entirely, is often the better approach. However, with lower-density plantings, trees are spaced further apart. Increasing the threshold can help more confidently group multiple attempts and improve filtering of failed strikes.

It’s also important to consider GPS accuracy in the background. For a standard STA Explorer, horizontal accuracy is typically around 1.06 m CEP (see our accuracy blog for more detail). When planting spacing approaches that level, spatial filtering becomes less reliable, which is why high-density scenarios often benefit from minimal or no distance-based filtering.

Thresholds are configurable on a per-customer basis, but currently need to be adjusted by our support team.

What’s coming next: adding a time threshold

In our next software update, we’re adding a time-based threshold alongside the existing spatial one.

This means that if multiple planting events occur in very rapid succession, they can be automatically classified as failed attempts, even if they fall outside the distance threshold. This adds another layer of confidence, particularly in situations where GPS variability makes spatial filtering less reliable.

Like the distance threshold, this will be configurable to suit different workflows and planting conditions.

Designed for real-world variability

There’s no single rule that works across all planting programs. Soil conditions, terrain, tool type, planting density, and operator behaviour all influence how events are recorded.

The STA platform is designed to be flexible, allowing thresholds and logic to be tuned so you get the most accurate representation of what’s actually happening on the ground.