AUTOMATING

VISUAL

INSPECTION

OF OFFSHORE

FACILITIES

Visual inspection is an essential component of assets and facilities management in the energy sector. In this demo, the SoftServe Energy team presents a step-by-step vision on how to build an end-to-end AI solution for visual inspection—relevant for any inspection application: corrosion, coating breakdown, pipelines leaks detection, cracks, and structural and equipment defects.

AREAS OF INSPECTION

VENT STACK & EXHAUST

INSPECTION

TANK

INSPECTION

STRUCTURAL

INSPECTION

UNDER DECK

INSPECTION

CHALLENGES

HIGH COST

OF QUALIFIED EXPERTS

Highly qualified inspectors cost a lot, as becoming one takes much effort, training, and experience. There also are additional costs related to manual inspection of offshore platforms, like worker's transportation, accommodation, equipment supplies, etc.

AMOUNT OF DATA & TIME FRAMES

 

 

Depending on the scope and area of inspection, data can be collected in large amounts. All of the data needs to be inspected and correctly interpreted—prior to any reporting and activities— leading to equipment downtime and additional expenses.

HEALTH & SAFETY HAZARDS

Visual inspection of offshore platforms is associated with high-risk activities—often requiring a person to access areas in high altitude, confined space, and other hazardous places. The actual location of platforms makes safety hazards issues worse, as they are situated far from an oceanic coastline, in different climate zones, and under evolving weather conditions.

HUMAN

ERRORS & LIMITATIONS


The mistake and bias in human judgment can lead to varying conclusions by different experts. Lack of specific training, qualification, and in some cases experience can lead to wrong interpretation and conclusions about equipment condition. Also, human eyesight can be imprecise and fooled by optical illusions, resulting in incorrect identification or missing of defects.

HOW IT WORKS

Step 1

3D MAPPING

Create a 3D map of a site (interior and exterior)

Step 2

IMAGE/VIDEO ACQUISITION

AND LOCALIZATION

Use UAVs, ROVs or other remotely controlled or autonomous unmanned vehicles inspect the site, collect and localize the media on a 3d map

Step 3

IMAGE/VIDEO ENHANCEMENT

Using specific AI/ML models to enhance images (Reducing blurriness, improving lighting, etc.)