AyniCam
How AyniCam documentation distinguishes facts from tests
Understand the evidence behind AyniCam documentation: release records, simulator captures, device tests and the limits of each type of proof.
Four kinds of evidence
A published App Store description establishes what a release advertises. Matching release-build records establish documented configuration and implementation. A simulator capture shows an interface under disclosed conditions. A physical-device test measures a particular device and scenario. None should be silently substituted for another.
What is verified in this documentation
The current product reference is version 2.0 build 6. Public release information, current App Store metadata and matching recording-source fingerprints support the described clip configuration. Existing software checks support specific release fixes. These checks do not establish a road-detection accuracy percentage, battery life, thermal endurance or behavior on every iPhone.
Simulator images and illustrative diagrams
Captures are labeled when they use simulator locations, weather, objects or camera fixtures. Feature illustrations explain concepts; they are not app screenshots. A visually convincing simulated scene is not evidence of real-world distance accuracy. Published figures should identify their origin and the limitation next to the image.
Requirements for a future device test
Record the iPhone model, iOS and app versions, battery health, permissions, model/view, recording plan, brightness, power source, ambient conditions and test duration. Use a safe stationary procedure or safely collected footage. Repeat the same procedure and preserve source files. For interruptions, compare the observed camera state, warning and actual saved output; for battery and heat, report conditions and repeated measurements.
Requirements for an AI evaluation
Define the labeled sample, supported classes, mistakes being counted and conditions excluded. Show missed objects as well as incorrect detections. A confidence label is not a measured collision risk. Report sample size and uncertainty and avoid converting a few attractive frames into a universal accuracy claim. Do not expose private routes, identifiable people or proprietary model files to make a test look more complete.
Publication and corrections
New documentation is published under AyniCam’s editorial responsibility. No personal founder review, independent certification or native-language review is claimed unless it has occurred. Physical-device background, night, battery and multi-device studies are not yet published here. When a result becomes available, its article should link its method, data scope, date and corrections.