Field lab · GPS analysis
When a tracklayer stops to place an article, the person stands still but individual GPS fixes continue to move. Knurrix therefore separates the untouched raw track from a cleaned map view and tests that logic against anonymised field recordings.
In one field test, 34 fixes clustered during roughly 30 seconds at a standstill. The map view reduced them to two representative points while retaining every original fix for inspection.
On one anonymised track, the strongest artificial bend in a curve fell from 66 to 26 degrees. The maximum display deviation from the raw points fell from 1.68 to 0.84 metres, while distance changed only from 454.4 to 454.0 metres.
On one real curve, 71 of 75 search points lay on the same side of the laid line. This can reveal a systematic tracking-line offset and is not automatically a dog error. The map supplies observations; expert assessment remains with the handler.
Frequently asked questions
No. The raw data is retained. The app creates an additional cleaned map representation and can calculate derived values from it.
Satellite geometry, obstruction, reflections and the device estimate change position fixes even without real movement. This is normal for GNSS and varies by environment.
No. A cleaned display remains a technical approximation. Knurrix therefore states the method and limitations and keeps the original track separately.
Help shape Knurrix
We would value your honest experience from the field. Register your interest for news and future testing rounds – with no obligation.