Infrastructure Change Detection in GIS. Change point detection (PIPELINE Monitoring)

Pipeline monitoring (infrastructure control) is performed by a combined remote sensing method: satellite imagery and UAV survey in daytime and at night.

Prompt detection: failures on oil-field pipelines; spills of oil, oil products and produced water; unauthorised taps and third-party actions; fires in forests and peatlands along pipeline routes; litter along the right-of-way; unauthorised storage of construction materials and pipes; unauthorised persons and vehicles in protection zones.

Condition assessment: aerial crossings and crossings of water barriers and ravines; valve assemblies, pig launcher and receiver units; along-route roads, bridges, crossings, culverts; off-design position of sections; repair works; kilometre posts and signs.

Control: violations and damage; sites of construction and installation works and earthworks; crossings and parking of machinery without documents agreed with the operator.

Automatic monitoring of power lines and the MDZ is automatic inspection of power lines from satellite and UAV images: line measurement, 3D reconstruction of the terrain along the corridor (MDZ) and obstacle recognition.

Machine-learning methods make it possible to track asset reliability and automatically detect overgrown vegetation, illegal construction in the right-of-way and natural-hazard events (landslides and floods).

The source separately notes: vegetation detection in power-line corridors from optics, SAR and LiDAR; new visualisation methods, including UAVs and autonomous vehicles; detection of hazardous conditions around critical assets.

Monitoring is carried out on a geospatial base from orthorectified images of very high spatial resolution (down to 1 m at nadir).

First, source data are collected: the area of interest, terms of reference, selection of archival satellite imagery or an order for a new survey. Then photogrammetrists process the images and prepare the base. Specialists analyse surrounding objects that fall into the MDZ. At the final stage they perform geoprocessing, quantitative assessments and enter into the database HPF trunk-pipeline sections with MDZ violations, linear coordinates, length and shortest distances to violators.

For power lines the source lists: detection of vegetation and its height; inspection of towers; search for breaks and damage assessment; data for prompt emergency response. Automatic detection of vegetation in the power-line protection zone, measurement of vegetation height, detection of damaged towers and broken wires.

The protection zone of highways in the Russian Federation is described in the source as having a vague value and ranging from 25 to 50 metres, and even less when passing through settlements. For railways the FAQ gives examples: at least 50 m to roads, parking lots, warehouses, public utilities and garden plots; at least 100 m — a fence, house, garage. Actual protection territories depend on location, number of tracks, route parameters and power-line voltage.

The remote sensing method is faster than field work, more convenient in complex natural-climatic and geological conditions, more effective for long and hard-to-reach hazardous facilities (trunk pipelines, power lines) and more acceptable in price/quality ratio.

Early detection reduces the risk of fines and fire-hazard situations with damage to property or a threat to life. According to the source, some sections of especially hazardous facilities fall into a high-risk group of man-made accident (with MDZ violation).

For roads the source also lists: a continuously operating MDZ GIS; environmental conditions on thematic maps; emissions and noise; recording of littering and illegal construction; NDVI and NDMI in MDZ buffers. For railways: the NDReLI index for corridor erosion, monitoring of landslides, ballast without vegetation, construction and reconstruction of infrastructure.