VISUAL INTELLIGENCE

Solera’s Artificial Intelligence-powered automated damage detection technology

Using the latest advances in image processing and deep-learning algorithms, our multi-model approach to damage classification and instance segmentation ensures image quality and pinpoints damage type and location within seconds, allowing users to “see what AI sees”.

30 seconds
to create an
estimate from photos

85%
F1 score

2000+
vehicle models
detected by AI

STATE-OF-THE-ART AI

UNMATCHED ACCURACY

Our machine learning models are trained by an unparalleled global database with more than 100 million images to identify the type and severity of damages with unmatched speed and accuracy.

MARKET AGNOSTIC

Our approach to parts and damage detection works for every body shape, regardless the vehicle make and model.

BUILT TO SCALE

Trained and served using Google Cloud technologies, our state-of-the-art and elastic AI infrastructure allows our data science to easily enhance and scale our solutions.

SUB-SECOND PERFORMANCE

Highly efficient setup GPUs (Graphics Processing Units) reduces processing times to under 1 second per image, allowing full estimates to be created in just under 30 seconds.

REAL-TIME IMAGE QUALITY FEEDBACK

Too dark, blurred, over exposed, too small or too large images are flagged. Our photo analytics algorithms detect poor image quality, generate warnings and request retake in real-time.

CONTINUOUS LEARNING

A highly automated data ingestion process and “human-in-the-loop” feedback loop ensures our machine learning models learn faster, deliver better predictions and fulfil new use cases.