Linker Vision ← All products / Electricity

Electricity

Transmission Tower Patrol

Autonomous drone patrol turns high-voltage transmission towers and substations into a self-inspecting asset base. Drones fly pre-planned routes from a docking station and capture RGB and infrared imagery; vision models locate bird nests, tower-structure corrosion and damaged insulators, while thermal readings are tied to the component they belong to instead of being read as a picture. Every finding is graded, geo-referenced to the tower, and written into an inspection database for trend analysis and replacement prediction. Phase 1 scope in the Kaohsiung Lighthouse extension covers about 100 high-voltage towers on fully autonomous drone routes, with substation bird-nest and thermal inspection in the same pipeline.

VLM output

A small quadcopter drone flies from the left side of the frame across a lattice transmission tower that carries mounted antenna panels, passing close to the structure and continuing to the right against a clear blue sky. The tower stands in dry grassland with overhead power lines and a residential neighborhood in the background; no people or vehicles are active in the scene.

  • Tower VisibleYes
  • Drone Inspection ActiveYes
  • Drone Count1
  • Tower Structural DamageNo
  • Vegetation EncroachmentLow
  • Power Line ObstructionNo
  • Human PresenceNo
  • VisibilityClear
  • Time Of DayDaytime

Customer challenges

1

Tower climbing and telescope inspection are manual and slow, and expose crews to fall and electrical risk

2

Drone footage is captured but graded by eye afterwards, so defects are missed and never scored consistently

3

Corrosion, insulator cracks and bird nests look like structural clutter in imagery; generic CV cannot separate them

4

Thermal images are read as pictures rather than as measurements attached to a specific component

5

Findings stay in inspection reports, so there is no trend, no severity history and no life-cycle prediction

Customer benefits

1. 15x inspection efficiency compared with manual tower inspection

2. Annual labour cost for one operating zone reduced by about US$13K-23K

3. Occupational-safety incidents down 70% by removing routine tower climbing

4. AI records every inspection result and notifies anomalies automatically, with no reliance on the inspector's notes

5. Findings flow into a database for long-term observation, trend analysis and prediction of component damage and replacement

Key features

Bird Nest Detection on Tower Structure (Drone Imagery)

Detects bird nests on tower structures from drone imagery.

Thermal Anomaly Detection (Infrared Temperature Inspection)

Detects thermal anomalies through infrared temperature inspection.

Structural Corrosion & Insulator Damage Detection

Detects structural corrosion and insulator damage.

Before & after

Before 1. Inspection efficiency at manual climb-and-look baseline, 1x After — with Linker AI Nexa 1. 15x inspection efficiency
Before 2. Tower coverage per inspection cycle, sampled subset After — with Linker AI Nexa 2. 100% of towers in scope, every cycle
Before 3. Occupational-safety incidents in inspection work at baseline After — with Linker AI Nexa 3. -70%
Before 4. Annual inspection labour cost per operating zone at baseline After — with Linker AI Nexa 4. About US$13K-23K per zone per year lower
Before 5. Defect grading consistency, inspector-dependent and unmeasured After — with Linker AI Nexa 5. Uniform AI grading on 100% of captured frames
Before 6. Share of findings usable for trend and replacement prediction, 0% After — with Linker AI Nexa 6. 100%, every finding graded and queryable