What you will learn
- What site investigation records and ground models contain
- How soil parameters are derived and recorded with uncertainty
- What slope stability, settlement, and bearing analyses require
- How foundation design and reporting use the ground model
- How to phase a tool from investigation records to full analysis
What geotechnical analysis software is
Geotechnical software builds the ground model from investigation data and uses it for design. Borehole logs, laboratory results, and in-situ tests are recorded and interpreted into soil layers and parameters; those parameters feed slope stability, settlement, and bearing capacity analyses; and the analyses support foundation and earthworks design. The output is a design supported by a traceable chain from investigation to recommendation.
Uncertainty is the honest centre of geotechnical work. Ground conditions vary between boreholes, parameters are interpreted rather than measured directly, and the software should present ranges, sensitivity, and the basis for each parameter, rather than a single confident value.
Core features
Traceable interpretation with uncertainty shown defines the product.
- Borehole and investigation records with logs and test results
- Ground model building: layers, parameters, and cross-sections
- Parameter derivation with basis, ranges, and sensitivity
- Slope stability, settlement, and bearing capacity analyses
- Foundation design modules linked to the ground model
- Reporting: investigation summaries, analyses, and recommendations
How it is built
Start with investigation records and ground models, then add analyses.
- Records: borehole logs, lab results, and in-situ tests for one project type
- Ground model: layers, parameters, and cross-sections
- Analyses: stability, settlement, and bearing capacity
- Design: foundation modules linked to analyses
- Reporting: summaries, recommendations, and revision control
Technology choices
The ground model should be a structured data model, not a set of drawings. Structured layers and parameters make analyses repeatable and let new investigation data update the model with visible consequences.
Analysis methods should be documented per project, with method references and version control, so a design can be reviewed years later against the method actually used.
Challenges to plan for
Interpretation disagreement is normal. Different engineers interpret the same boreholes differently, and the system should support alternative interpretations and sensitivity views, rather than forcing one interpretation early.
Laboratory data quality varies. The system should flag incomplete, inconsistent, or outlier results for review, instead of silently feeding them into parameters.
Who it is for, and its limits
Geotechnical software fits geotechnical consultancies, civil engineering firms with geotechnical teams, and infrastructure projects with significant ground risk. It fits projects where ground conditions drive design.
The software does not investigate ground. Site investigation, sampling, and engineering judgement stay with geotechnical engineers; the tool records, models, and analyses their work.
What a first release looks like
A first release should record boreholes and laboratory results and build ground models with layers and parameters for one project type. Stability and foundation analyses come later, once ground models have supported real design decisions, including a model updated by new investigation data with the consequences visible. Parameter uncertainty must be shown from the start, because single-value parameters hide the interpretation risk that matters most in geotechnical design.
Consider a site where two engineers interpret the same boreholes differently on one layer boundary. The system holds both interpretations, shows the sensitivity of the foundation recommendation to each, and the project engineer selects one with recorded reasoning, while the alternative remains visible for review. New laboratory results arrive mid-design, the parameters update with versioning, and the affected analyses are flagged for rechecking. If the model can only hold one interpretation, the engineering judgement is being hidden, not recorded.
A mistake worth avoiding
The mistake worth avoiding is designing from a single ground interpretation as if the ground were known. Boreholes sample a tiny fraction of a site, and the ground between them is interpreted, not observed; a design that only works under one interpretation fails when the ground proves different. The discipline is sensitivity-based design: alternative interpretations modelled, design sensitivity to parameter ranges tested, and contingency — additional investigation, observational methods, conservative parameters — chosen where sensitivity is high. A geotechnical design that cannot say what happens if the ground differs from the model is not a design; it is a hope.
Frequently asked questions
What is a ground model?
An interpreted model of soil and rock layers and their parameters, built from borehole logs, lab results, and in-situ tests, with cross-sections showing how layers vary across a site.
Why show parameter ranges?
Because soil parameters are interpreted from limited data. Ranges and sensitivity views show how much design conclusions depend on uncertain values.
How are analyses checked?
Through documented methods, versioned inputs, and independent review of the analysis pack, like structural calculation checking.
Can new investigation data update the model?
Yes, with versioning: the model, analyses, and design consequences of new data should be visible, not silently overwritten.
How much does it cost to build?
Cost depends on analysis types, project scale, and reporting depth. Investigation records and ground models alone are far smaller than a full analysis suite, so confirm with a scoped quote.