Reviewers have always judged coursework photographs by looking at them. Provenance replaces that guess with an audit of what the camera itself recorded: the sensor's timestamp, the coordinates, the hardware signature, and a hash that no rename survives.
0
images uploaded
Analysis is local
4
objective checks
Per submission
256
bit digest
Rename-proof
1
shared ledger
Across all students
The image is the claim. The metadata underneath it is the evidence. Scrolling separates one from the other.
↓ Scroll down to decompose evidence
Reads the original capture timestamp straight off the camera's sensor clock and tests it against the scheduled practical window.
Derives an immutable 256-bit hash. Server-authoritative matching flags a file resubmitted under any name, by any student.
Extracts latitude, longitude and altitude, then reverse-geocodes them against the faculty's laboratory and fieldwork geofences.
Cryptographically audited evidence: zero subjective guesswork.
Original EXIF telemetry intact. Coordinates and capture window both match the assigned field exercise.
Verification matrix
Extracted EXIF telemetry
Computed in-browser · WebCrypto
Cryptographically auditedCapture timestamp
Reads the original EXIF timestamp written by the camera's own clock, and tests whether the photograph was taken inside the scheduled practical window rather than recycled from a previous year.
tags → DateTimeOriginal
GPS telemetry
Decodes the latitude, longitude and altitude embedded in the raw file — from the EXIF GPS block, an XMP packet, or a PNG's eXIf chunk. Where a file carries none, the specimen can instead be photographed inside Provenance, which reads the device position as the shutter fires.
tags → GPSLatitude · GPSLongitude · eXIf · XMP
Optical signature
Extracts the sensor make, lens profile, focal length and ISO. Their absence is itself evidence: editing software, screenshots and messaging apps all leave the same gap.
tags → Make · Model · FocalLength
Duplicate detection
Hashes the file to a 256-bit digest and matches it against the departmental ledger, so an identical photograph submitted twice is caught across students, not merely within one.
tags → SHA-256 · crypto.subtle
Coursework photographs carry research specimens, unpublished lab setups and the faces of the people who did the work. So the file stays where it was taken. Parsing and hashing run in the browser through exifr and crypto.subtle.digest; what is filed is the reading, not the photograph.
The server is still authoritative for verdicts — only it can see every student's digests. That limitation is documented, not hidden.
Buffered on device
The original camera file is read into memory. It never leaves.
Binary EXIF parse
Timestamp, GPS and hardware tags decoded locally by exifr.
Verdict derived
Four rules evaluated, then re-derived server-side against the ledger.
Certificate issued
An archival audit sheet, printable for moderation and grading.
Everything that reads the image runs inside your browser: exifr parses the file's EXIF segment, and crypto.subtle.digest computes the SHA-256 hash. The full-resolution photograph is never transmitted. What gets filed is the derived record — the digest, the extracted values, the verdict — together with a thumbnail no larger than 96 pixels, so a reviewer can see what was submitted.
Physics
PHY
Geology
GLY
Chemistry
ICH
A verdict in under a tenth of a second, computed on your own device, with nothing transmitted but the reading.