> For the complete documentation index, see [llms.txt](https://docs.paveapi.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.paveapi.com/how-pave-works/accuracy-and-limitations.md). # Accuracy & Limitations PAVE is designed to match the accuracy of in-person inspections, despite some limitations in damage detection from photos alone. PAVE's grading capabilities have been extensively tested and improved over the past two years with the help of one of North America's largest auto auctions. In April 2022, the most recent test was conducted, comparing the PAVE reports for 200 vehicles to the in-person inspection reports from the auction. ## PAVE Damage Detection Limitations ### Interior Inspections Not Included Our User Acceptance Testing has shown that users are less likely to complete the inspection process if it requires more than 13 photos and takes longer than 3 minutes. As a result, a complete interior inspection using PAVE's approach is not feasible. However, the grading algorithm accounts for the interior to be in similar condition to the exterior. Tests have shown that when PAVE grades a vehicle as being in Poor Condition, it is expected that the interior will also be in poor condition. ### Glass Damage May Not Show in Photos PAVE inspects all glass surfaces for cracks, chips, scratches, and stars. On average, 70% of these types of damages can be seen in the photos captured. However, small chips and stars in a windshield may not be visible in the photos and therefore, undetectable by PAVE. ### Previous Accident Repairs Detection of previous paintwork is best done with a paint gauge to measure the thickness of paint on each panel. PAVE will detect mismatched paintwork and misaligned panels, but it cannot effectively detect previous accidents. We recommend checking the vehicle's history report to confirm any reported accidents. ## --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.paveapi.com/how-pave-works/accuracy-and-limitations.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.