Corrosion and insulation damage are notorious culprits behind unplanned downtime, high maintenance costs and safety risks. For industrial asset owners, early detection is therefore essential. Thanks to a new AI-driven inspection service by Smart Asset Integrity Solutions (Smart AIS) and Airvision, this process is now significantly more efficient. This joint service combines practical field expertise with advanced image recognition and artificial intelligence – making inspections faster, safer and more reliable.
The service begins with a site survey: Smart AIS maps the installation using drones and/or 3D laser scanners. An AI model then analyses the visual data, identifying potential areas of concern, such as corrosion or insulation damage. This involves applying an AI algorithm to the collected imagery.
Prioritising maintenance
Each marked spot is assigned a severity score (from 1 to 5) by Smart AIS, helping asset owners determine maintenance priorities. Typically, a score of 5 prompts immediate targeted repairs. Asset owners can also use the AI’s pre-selection to guide their teams in conducting follow-up inspections, without having to manually check every asset first.
Technology meets expertise
Although technology greatly accelerates the inspection process, human expertise remains vital. The algorithm detects damage with over 90% accuracy, after which a human inspector reviews and validates each finding. This creates a smart blend of automation and craftsmanship: the system takes over the bulk of the work, while inspectors focus on interpretation and final decision-making.
faster, safer, more accurate
The benefits are significant. Traditional visual inspections are time-consuming and physically demanding. By combining drone footage, 3D laser scans and AI analysis, this new method is three to five times faster. Safety is also improved: since inspectors no longer need to physically access all areas, there's less need for scaffolding or rope access – reducing both risk and cost.
collaboration with airvision
The strength of this service lies in collaboration. Smart AIS contributes years of field experience, along with a large library of annotated inspection imagery. This dataset has been used to extensively train the AI model. Airvision, a specialist in image recognition with AI, helped develop a system tailored to the needs of industrial clients.
designed for ease of use
Ease of use was a core focus during development. The AI platform has been aligned step by step with the practical workflows of inspectors. Results can easily be exported as Excel files, PDF reports or datasets for integration with asset management systems like SAP or Maximo. This paves the way for a fully digital maintenance chain – where inspection results directly generate repair orders.
This initiative is part of a broader digital inspection chain, in which other solutions of Smart AIS and partners also play a role. Read more about the full inspection chain here.
applicable and scalable
The technology has already been successfully deployed at Huntsman Holland in Rotterdam, where it proved to be not only faster and safer but also more reliable than traditional inspections. Other asset owners – such as those at the Chemelot site – have also shown interest. While the service is currently aimed at industrial environments, it is broadly applicable. The AI model currently focuses on detecting corrosion and insulation damage, but the system is easily expandable – for example, to spot missing fire extinguishers, loose parts, thermal anomalies or leaks. With enough data, new applications can be developed quickly.
ready to get started
Companies that already have inspection data can start using the system immediately. Those without images can request an initial scan from Smart AIS to create a usable dataset. Depending on their needs, clients can either process the results themselves or outsource the entire project to Smart AIS.
For more information about the AI-based inspection service, please contact Freek Lovink.
For more information about Smart AIS: smart-ais.eu
For more information about Airvision: airvision.ai