Belt and Road Infrastructure Project Progress Tracking

Monitor large-scale infrastructure investments across Central Asia, Africa, and Southeast Asia. Klarety agents run Sentinel-2 change detection time-series to flag new road alignments, railway embankments, port expansions, and industrial zone clearing. One prompt returns a project-level completion status report with delay risk flags.

View projects

Infrastructure completion tracking across continents

Klarety AI

Project completion scores from satellite data

Klarety agents run change detection on Sentinel-2 and return project-level completion scores with delay risk flags.

Sentinel-2

Road alignment and port expansion detection

Agents classify new road alignments, railway embankments, and port expansions from spectral change signatures in time-series.

GIS Output

Infrastructure vectors for investment analyst GIS

Export road and rail alignment vectors and construction polygon shapefiles for project finance GIS tool integration.

Change detection for infrastructure completion scoring

Klarety agents apply NDVI-loss and SWIR-change detection to Sentinel-2 annual mosaics to map road and rail corridor clearing, grading, and surfacing progression. Completion scores are computed as the fraction of planned corridor length showing construction signatures. Output is a segment-level completion table and corridor shapefile with delay risk flags.

Klarety AITrack Belt and Road progress in your region

analysis/bri_road_progress.pyAgent code

python

import eeimport geopandas as gpdimport numpy as npfrom shapely.geometry import LineString
ee.Initialize()
# Example: Central Asian BRI corridor segmentcorridor_aoi = ee.Geometry.Rectangle([69.0, 40.5, 76.0, 42.5])
# Annual mosaics: baseline vs currentdef annual_mosaic(year):    return (ee.ImageCollection('COPERNICUS/S2_SR_HARMONIZED')            .filterBounds(corridor_aoi)            .filterDate(f'{year}-01-01', f'{year}-12-31')            .filter(ee.Filter.lt('CLOUDY_PIXEL_PERCENTAGE', 20))            .median()            .select(['B4','B8','B11']))
before = annual_mosaic(2022)after  = annual_mosaic(2025)
# NDVI loss (vegetation removal = clearing for road/rail)ndvi_b = before.normalizedDifference(['B8','B4'])ndvi_a = after.normalizedDifference(['B8','B4'])ndvi_loss = ndvi_b.subtract(ndvi_a).rename('ndvi_loss')
# SWIR increase (bare soil, gravel, asphalt)swir_delta = after.select('B11').subtract(before.select('B11')).rename('swir_delta')
# Infrastructure mask: NDVI loss + SWIR increaseinfra_mask = ndvi_loss.gt(0.12).And(swir_delta.gt(600))
# Estimate linear km of new constructioninfra_area = infra_mask.multiply(ee.Image.pixelArea()).reduceRegion(    ee.Reducer.sum(), corridor_aoi, 10, maxPixels=1e9).get('ndvi_loss').getInfo()
# Assuming 30m road width: area / 30m = linear metreslinear_km = (infra_area or 0) / 30 / 1000
print(f"Infrastructure change area: {(infra_area or 0)/1e6:.2f} km2")print(f"Estimated new construction: {linear_km:.1f} linear km")print(f"Planned corridor: 420 km")print(f"Completion estimate: {min(linear_km/420*100, 100):.1f}%")

output/bri_progress_report.mdOutput report