The geotechnical characterisation of silty soils remains a challenge due to their transitional behaviour and high variability. In Finland, silts are widespread but poorly represented in existing correlations, which are largely based on clays or sands. Current practice often relies on Weight Sounding (Painokairaus, PK) and Combined Static–Dynamic Penetration Testing (Puristinheijari, PH), supported by empirical guidelines developed several decades ago. While these methods are cost-effective, their reliability in silts is uncertain and often conservative. In this study, we reported results from a benchmarking campaign at the Haistila test site in south-west Finland, where PK, PH, and piezocone penetration tests (CPTU) were performed independently by Tampere University and Ramboll Finland Oy. The objective was to quantify variability between methods and operators, and to assess implications for geotechnical design. Results showed that CPTU provided the most repeatable measurements, with cone tip resistance showing the lowest relative error and coefficient of variation. In contrast, PH results displayed greater variability, particularly in torque, and resulted in possibly conservative design parameters according to national guidelines. The findings confirmed that PK and PH are useful for stratigraphic profiling but not for parameter derivation in silts. CPTU, if calibrated with laboratory data, offers a more robust alternative and highlights the need for updated, silt-specific correlations in Finland.

Peat is a highly variable organic soil that presents major challenges for geotechnical engineering. In Finland, where peatlands cover one-third of the land area, infrastructure often intersects deep deposits that are difficult to characterise and costly to improve. Conventional practice relies on conservative assumptions or large-scale replacement, which can be expensive and carbon intensive. In this study, we compiled a harmonised database of over 250 datapoints from Finnish, Nordic, and Western European sources, focusing on compressibility, yield stress, and undrained shear strength. Regression analyses were benchmarked against Random Forest machine learning models. The results confirmed that the compression index is well predicted from water content, whereas yield stress and undrained shear strength display high variability under regression. Random Forest models provided modest improvements over conventional regression for strength-related parameters, while compressibility remained well captured by empirical correlations. Cross-parameter estimators offer additional tools where direct strength testing is unavailable. The findings underline the complementary roles of regression and machine learning in peat characterisation. Moreover, a hybrid workflow is proposed: regressions for early screening and conservative design, and machine learning for refined, site-specific assessment, supporting more sustainable infrastructure development on peatlands.