An evidence trail.
Bring observations, inspection records and dated interventions into a traceable history for each study segment.
EARTH OBSERVATION / RIGHT-OF-WAY
SPACE + TIME
ROW4D
Satellite observations for right-of-way vegetation management.
Research and validation pilots
Every stretchhas its time.
Follow your corridor through successive satellite observations. A ROW4D pilot connects this history with field evidence and dated interventions to test management priorities, stretch by stretch.
For transmission maintenance, environmental, asset and R&D teams.
FROM ORBIT TO THE SAME STRETCH
A pass. An observation.
The satellite observes an area of the corridor. Each observation belongs to a place and a date.
THE DECISION COMES FIRST
01 / 04Plan fromwhat changes.
The pilot starts with the decision your team needs to make. Together, we define the corridor, the management threshold and the evidence needed to test the approach.
The same place. Seen again.
Successive satellite observations record the corridor on different dates. Compare the same stretch over time to identify changes that need investigation.
A closer look at the same stretch.
Dated inspections help interpret vegetation type, height and local conditions.
What happened between observations.
Recorded mowing and pruning give the vegetation timeline a reference point.
Working outputs, defined during scoping
Priorities with context.
Review proposed priorities alongside the evidence that supports them and the uncertainty that remains.
A comparison your team can assess.
Test the proposal against the current routine, using criteria agreed before the pilot starts.
ONE CORRIDOR. TWO QUESTIONS.
02 / 04One corridor.Two questions.
A
VEGETATION / REGROWTH
What grows back.
Follow regrowth after mowing or pruning and investigate when it approaches the management threshold agreed for the corridor.
B
FUEL CONDITION
What can fuel a fire.
Investigate how vegetation and environmental context change the fuel available along the right-of-way.
Each workstream has its own variables, field references and success criteria.
A PILOT BUILT AROUND YOUR CORRIDOR
03 / 04Start witha real stretch.
Choose a management question that matters to your team. Build a test around it.
- 01
Define the decision.
Select a corridor and the management question your team needs to answer.
- 02
Bring the records.
Relate dated satellite observations to field inspections and intervention history.
- 03
Agree on the test.
Set the outputs, comparator and success criteria before the pilot starts.
Scope and timing follow the available data and the decision being tested.
Operational decisions remain with the responsible team. Field surveys and measured geometry are included where the question requires them.
THE RESEARCH RECORD
Evidence shapesthe next test.
The research has a documented history. Its findings inform how the next pilot should be designed.
Read the research record
An initial historical baseline
The initial MapBiomas Fire C5 extraction covered 1985–2023 across 4,618 study segments in the BMTE and XRTE corridors.
A later retrospective experiment
In the burned-area benchmark, the evaluated model did not outperform the strongest comparator at the main horizon in either held-out year.
| Held-out year | Evaluated model M1 | Strongest comparator |
|---|---|---|
| 2024 | 25.50% | 29.46% |
| 2025 | 25.50% | 32.38% |
Recall@10% measures the share of positives captured among the highest-scored 10% of segments in each issue. H60 is two calendar months in this experiment.
These results guide the next test design. Estimating time to a vegetation-management threshold requires separate validation.
The experiment concerns mapped burned area along the right-of-way. It does not establish operational ignition forecasting, electrical-fault prediction or a validated intervention date.
Research records
ROW4D final experiment report, 14 October 2026.
ROW-4D M0 closeout report, 6 October 2026.
The 14 October report supersedes the experimental status described on 6 October.
Before we scope a pilot.
What does ROW4D mean?
ROW stands for right-of-way. 4D expresses the project’s spatial and temporal perspective: understanding a corridor through successive observations. Resolution and accuracy depend on the data available for each pilot.
Which records help us start?
Georeferenced corridors or spans, dated mowing and pruning records, and field inspections with vegetation type, height and photographs. Questions involving proximity to conductors also require measured cable geometry.
What is a management threshold?
A condition agreed with the operator for assessing the need for intervention. The threshold is defined for the management decision and validated against the appropriate field evidence. Electrical-clearance questions require their own measurements and validation.
How is the pilot assessed?
Outputs, comparator and success criteria are agreed before execution. The evaluation considers the decision being tested, the quality of the evidence and the uncertainty in the result.
YOUR NEXT STRETCH
Tell us aboutyour corridor.
Bring us the management question. We will define what a useful pilot needs to test.
ricardo@grafter.com.br ↗