A comparison of lidar accuracy across iPhone models, with implications for reproducibility and cross-study comparison

iPhone lidar facilitates mobile handheld laser scanning, with applications spanning archaeology, engineering, and forestry, but hardware updates impact small object scanning. As older models are discontinued, it is imperative to ascertain how previous assumptions hold across newer lidar sensors, to safeguard reproducibility and cross-study comparison. Lidar accuracy was compared between the iPhone 12 Pro, iPhone 15 Pro and a commercial handheld scanner, the GeoSlam Zeb Horizon. Fifteen objects (0.8–42.5 cm) were scanned at 0.01–3.00 m. Root mean squared errors (RMSE) of overall point cloud distances to known validation primitive shapes were nearly identical between the iPhone 15 (2.06 cm) and the iPhone 12 Pro (2.05 cm). Direct cloud-to-cloud comparisons showed an RMSE of 0.90 cm between iPhones. X, Y, Z, dimension estimates were often slightly more accurate for the 12 Pro than the 15 Pro, likely due to point count differences by distance. The iPhone 15 Pro performed better when scanning small objects at close range, but worse at greater distances. Zeb Horizon RMSEs were consistently a tenth of iPhone Pro lidar. Human decisions (e.g. app choice, settings) impact iPhone lidar accuracy more than differences between models, with the similar accuracies of iPhone 12 Pro and 15 Pro implying consistency across devices.