Defect Detection
AI is increasingly being used to identify defects in buildings – particularly during the construction phase – to assist with quality control and management. By leveraging machine learning and computer vision algorithms, AI can analyse visual footage to assess the condition of a building and flag potential issues like cracks, damp, corrosion, or incomplete installation.
What these technologies are able to identify is different depending on where in the property lifecycle they are applied. There is growing use of AI during construction to analyse high-resolution images and videos for the detection of potential structural defects or deviations from the original design. This helps project managers address issues proactively, avoiding rework. Some tools can compare visual site data to BIM models or schedules in real time.
Within residential property, there is an increasing use of AI for basic property condition evaluations, especially during lower risk activities such as rental listings, mortgage pre-approvals, or initial condition checks. These systems are designed to score the quality of a property based on listing photos. Often used by letting agents, lenders, or online platforms, these tools are designed to quickly evaluate the condition of a property without a full survey. Though they certainly aren’t a substitute for professional inspections, AI helps to provide a fast, inexpensive way to identify potential issues at scale.
However, despite clear benefits, the accuracy of the output is dependent on the quality of the input, and poor lighting, low resolution or hidden views can distort the results of the assessment. Therefore, when making decisions, it is critical that AI supports – but does not replace – human, professional expertise. When using AI during property condition assessments, it is important to interpret the results and apply professional judgement to ensure that the output is accurate, reliable and appropriate.
Report Generation
Report writing is a typical part of building surveying activity. AI tools are being used to streamline this process by summarising information – such as measurements, notes, and images – to produce structured reports. Natural language processing (NLP) and large language model (LLM) techniques are used to analyse data collected during the survey, converting the raw inputs into structured, meaningful information. By automating this process, surveyors can reduce the time spent on preparing documentation and focus more on inspection and analysis.
However, there are risks involved with using AI tools for report generation. The reliability of the output is heavily dependent on the foundational data provided, and while surveyors are responsible for collecting this data, the way it is processed by the AI can be difficult to understand, as it is not always clear how the model prioritises or interprets the information.
It is also important to consider that reports generally contain financial, or third-party information. Therefore, it is essential to ensure that the AI system used complies with relevant data protection regulations, such as GDPR, and data is only uploaded to the AI tool in appropriate circumstances. When using proprietary tools, users must understand how and where data is stored and processed to ensure the safe handling of data.
To use AI-powered report generation tools responsibly, human, professional oversight remains essential. Surveyors must review and validate the AI-generated content to ensure it accurately reflects the site conditions, complies with professional standards, and considers any contextual factors that the system may have missed. This human check helps maintain quality, accountability, and trust in the final report.
Principal author: Aaliyah Pollock