Designed and field-validated, not just tested in a lab

GeoSol Studio was developed and tested on real sites in Senegal, co-developed with a geotechnical engineering firm.

3 pilot sites in Senegal

Co-developed with GEOMAT INGENIERIE (Dakar)

Pilot site: geolocated boreholes, 3D geological modeling and Leave-One-Out cross-validation.

Pilot site: geolocated boreholes, 3D geological modeling and Leave-One-Out cross-validation.

Full pilot site: boreholes, 3D model and all 5 laboratory tests (Atterberg, grain-size, oedometer, direct shear, swelling).

A quantified reliability, not a visual impression

1. Removal — each borehole is removed, one at a time, from the dataset used for kriging.
2. Prediction — the model re-estimates the value at the removed point, from the remaining boreholes.
3. Comparison — the predicted value is compared to the value actually measured in the field.
4. Indicators — repeating the operation across all boreholes produces objective indicators: RMSE, MAE, bias.

These indicators, verifiable by your client or a control body, are included directly in the final PDF report.

Why this scientific rigor matters

Reduced uncertainty and lower cost of soil studies.

Fewer incidents caused by poor knowledge of the subsurface.

Safer structures, protection of populations.

Optimized earthworks and excavated volumes.