Calum Maclean, Projects Director at Marine Technical Limits, examines the uncertainty involved in assessing ageing floating offshore structures. Engineers may have only a short opportunity every few years to inspect inside a vessel’s tanks, yet their findings must help inform decisions about maintenance, safety and continued operation.
An inspection can cover only part of a structure. Engineers must consider what they found, what they could not reach and what they still do not know about loads, degradation or how a tank has been used. Calum argues that integrity engineering depends on understanding the size of those gaps. AI can analyse the information collected, but it cannot make an uninspected area certain simply by producing a confident answer.
The presentation makes a practical case for recording more than defects. Teams should also document which planned inspection tasks were not completed and which areas were checked without finding a problem. Both kinds of information matter when planning future inspections and judging whether a new finding represents a change. Without that context, a record of defects alone gives an incomplete picture.
Calum also warns against mistaking a large volume of inspection material for useful evidence. Hours of video or photographs may add little if they are poorly labelled or fail to show the areas that matter. Organising observations around specific locations and tasks makes them easier to interpret and gives AI a stronger basis for finding connections across the data.
Marine Technical Limits has found AI useful for reviewing and comparing organised information, but Calum stresses the need to question its conclusions, especially when several uncertainties build on one another. Watch the full presentation to hear his offshore examples and explore how engineers can use AI without losing sight of what remains unknown.
Speaker
- Calum Maclean, Projects Director, Marine Technical Limits
