Utilising Artificial Intelligence (AI) can save time, including in inspection work, says Freek Lovink of Smart AIS. That is why Smart AIS opted for a strategic collaboration with the company Airvision. The initial results are promising. ‘AI is a “hot topic” that many companies are exploring’, says Freek. ‘I am particularly interested in how to turn data from various sources into useful information, and AI is set to play a significant role in that.’
‘At Smart AIS, we have pooled forces with Airvision’, says Engineer Freek Lovink. ‘We did this to elevate our AI capabilities. Through the collaboration, we combine our expertise in inspection work with Airvision’s advanced technologies. At Smart AIS, we focus on what we do best: asset integrity, 3D laser scanning, and inspection services including data collection. Airvision focuses on developing advanced AI technologies. This collaboration allows us to strengthen each other, leading to innovative solutions.’
Accurate and fast
The specific AI innovation we have developed together revolves around automating inspections through image recognition and machine learning. This is also very useful in industrial environments, Freek explains. ‘Smart AIS collects data during field inspections. This data consists of drone images, 3D scans, and handheld camera recordings. Our inspectors then manually analyse the collected data. In this phase, damages are detected by evaluating images within a digital inspection platform. We provide commentary on the detected damages, detailing the type of damage, such as corrosion on pipes, corrosion on insulation, or insulation damage, and a severity score. We then create a report based on the analyses, containing information about the inspected parts with damage. This way, Smart AIS converts inspection data into useful information – reports – for our clients.’
Added value in the analysis phase
Airvision is an inspection platform that enables Smart AIS to perform all the above steps, with an interesting addition: Artificial Intelligence. Freek: ‘The Airvision team is very skilled in AI; that is truly their expertise. The added value of integrating AI into our workflow is mainly in the analysis phase. This means that Smart AIS inspectors have to perform much fewer manual tasks, because the AI models can increasingly better predict the type and severity of the damage. By using Airvision, we expect our analysis phase to proceed faster because we can deploy well-trained AI inspection models.’
Training AI
‘We are currently training the AI system using data from existing, previously conducted Smart AIS inspection analyses’, Freek continues. ‘This way, we teach the AI models to recognise damage. Subsequently, we ensure continuous improvement of these results as our inspectors provide ongoing feedback on the AI predictions coming from the Airvision platform. It is a real win-win: Smart AIS shares its inspection knowledge, and Airvision uses it to develop a better inspection platform. Subsequently, we expect that by using self-learning AI models, we will increase our productivity and provide our clients with better and faster service.’
Benefits
For Smart AIS clients, working with AI offers several benefits: by using the latest AI techniques, they can work more efficiently, faster, and at lower costs. The AI system helps clients by automating repetitive tasks without compromising on quality or safety. This leads to faster inspections, better decision-making, and ultimately cost savings. Clients benefit from the most advanced technologies on the market, allowing them to optimise their maintenance processes and improve their operations.
For more information about Airvision: https://airvision.ai/
About Airvision
Airvision, a part of Datacation, is located at the High Tech Campus in Eindhoven and the VU Campus in Amsterdam. The company specialises in data analysis and AI solutions. They offer services such as machine learning, text mining, dashboarding, and data infrastructure. Their team of econometricians and data scientists works on improving business processes through advanced algorithms and big data analysis.