Your browser doesn't support javascript. This means that the content or functionality of our website will be limited or unavailable. If you need more information about Vinnova, please contact us.

SaddleScanner - for a more precise saddle fit

Reference number
Coordinator SaddleScanner Sweden AB
Funding from Vinnova SEK 321 664
Project duration November 2025 - July 2026
Status Completed
Venture Innovative Startups
Call Innovative Startups 2025

Important results from the project

The goals were partly met. A training dataset of some 400 scans covering around 70 saddles and fifty horses was built, against a target of 100 examples. A cloud data platform and an app for delivering the service were put into operation. Models for noise reduction and for saddle segmentation and orientation work. The software was demonstrated in real time to hundred visitors at Falsterbo Horse Show, generating some sixty interested parties. Full automatic saddle placement remains.

Expected long term effects

If saddle fit can be measured rather than judged subjectively, horses´ back health and working lives improve. Research shows nearly 90 per cent of horse-rider combinations have fit problems, while 95 per cent of owners believe the saddle fits. Digital matching also cuts transport of heavy trial saddles and poor purchases, extending saddle lifespan. Expertise that today rests with individuals can be formalised and made available where experienced fitters are scarce.

Approach and implementation

The project ran in three work packages: data collection, ML model development and launch. Data collection and the technical foundation followed the plan. Building the app that delivers the service required more resources than expected, and a key person left early, delaying the software. Two degree projects were linked to the project, contributing ML development and an external UX evaluation. Validation with saddle fitters was smaller in scope than planned and continues beyond the project.

External links

The project description has been provided by the project members themselves and the text has not been looked at by our editors.

Last updated 28 August 2026

Reference number 2025-01982