# geotechnical engineering

**By Dan Cumberland** · Published May 20, 2026 · Categories: AI Strategy

> At its core, geotechnical engineering studies how soil, rock, and groundwater behave under load— then uses that understanding to design foundations,...

## What Is Geotechnical Engineering?

At its core, geotechnical engineering studies how soil, rock, and groundwater behave under load— then uses that understanding to design foundations, earthworks, and earth\-retaining structures[1](/blog/blog-geotechnical-engineering#ref-1)\.  Where structural engineering asks "will the building hold?", geotechnical engineering asks "will the ground hold the building?"

The field rests on three earth materials, and each one adds a different problem to solve:

- **Soil**— loose, layered, and variable; it compresses, shifts, and loses strength when saturated\.
- **Rock**— stronger but fractured; its behavior depends on joints, faults, and weathering as much as the rock itself\.
- **Groundwater**— the wild card; water pressure changes how soil and rock carry load, and it rarely sits still\.

Two knowledge domains anchor the work: **soil mechanics** \(how soil deforms and fails\) and **rock mechanics** \(the same questions for rock masses\)\.  The purpose behind all of it is single\-minded— confirm that the ground can carry the structure safely, before anyone pours concrete\.  Karl Terzaghi, regarded as the father of modern soil mechanics, established that footing in 1925[1](/blog/blog-geotechnical-engineering#ref-1)\.

That definition raises an obvious question: how is this different from the civil or structural engineer on the same project?

## Geotechnical vs Civil vs Structural Engineering

Geotechnical engineering is a specialization within civil engineering, not a separate field— and it differs from structural engineering by focus: geotechnical engineers don't design the building, they make sure the ground beneath it can carry the load[1](/blog/blog-geotechnical-engineering#ref-1)\.  Geotechnical engineering is civil engineering applied to the earth itself\.

Civil engineering is the parent discipline\.  It covers roads, bridges, water systems, and the broad practice of building public infrastructure\.  Structural engineering, often working alongside it, designs the frame that carries the load— beams, columns, slabs, connections\.  The line between geotechnical and structural is the line between the ground and the thing standing on it\.

```html-table
<table><thead><tr><th>Discipline</th><th>Primary focus</th><th>Typical deliverable</th></tr></thead><tbody><tr><td>Geotechnical</td><td>The ground— soil, rock, groundwater behavior</td><td>Geotechnical report with foundation and earthwork recommendations</td></tr><tr><td>Civil (broad)</td><td>Infrastructure systems and site design</td><td>Site plans, grading, drainage, roadway design</td></tr><tr><td>Structural</td><td>The structure itself— frame and load path</td><td>Structural drawings and calculations for the building</td></tr></tbody></table>
```

On a real project, these three collaborate constantly\.  The geotechnical engineer sets the bearing capacity and foundation type; the structural engineer designs columns and footings to match; the civil engineer ties it into grading and drainage\.  The "blurry line" people notice is just specialization at work, not a turf war\.  Each role answers a different question about the same site\.

So what does that specialization actually produce on a project?  It runs a defined workflow\.

## What Geotechnical Engineers Actually Do

A geotechnical engineer investigates subsurface conditions, tests the soil and rock that come back, and turns that data into design recommendations for foundations, retaining walls, slopes, and excavations— all delivered in a geotechnical report[1](/blog/blog-geotechnical-engineering#ref-1)\.  Site investigation produces the data; testing quantifies it; the report converts it into design you can build to\.

The work runs as a sequence, and each step feeds the next:

1. **Site investigation\.** Engineers drill boreholes, dig trial pits, and collect in\-situ samples to map what's actually under the site— soil layers, rock depth, and the water table\.  You can't design for ground you haven't seen\.
2. **Testing\.** Recovered samples go to the lab while other tests run in the ground itself, quantifying strength, density, compressibility, and drainage behavior\.  This is where assumptions become numbers\.
3. **Design recommendations\.** The engineer translates those numbers into specifications— foundation type, allowable bearing pressure, settlement limits, earthwork criteria— and documents them in the geotechnical report[5](/blog/blog-geotechnical-engineering#ref-5)\.

Along the way, the engineer manages a short list of failure modes that have humbled plenty of projects: **settlement** \(the ground sinking unevenly under load\), **slope failure** \(soil sliding downhill\), **liquefaction** \(saturated soil behaving like a liquid during an earthquake\), and **bearing failure** \(the ground simply giving way\)[1](/blog/blog-geotechnical-engineering#ref-1)\.  Each one is a question the investigation is designed to answer before it becomes a problem on site\.

None of this is guesswork\.  It's a disciplined chain from observation to data to design— and skipping any link weakens the whole thing\.  Each of those steps draws on a few core knowledge domains\.

## The Core Sub\-Disciplines of Geotechnical Engineering

Geotechnical engineering rests on three core sub\-disciplines: soil mechanics \(how soil deforms and fails under stress\), rock mechanics \(the same for rock masses\), and foundation engineering \(designing the structures that transfer load into the ground\)[1](/blog/blog-geotechnical-engineering#ref-1)\.  Soil mechanics is the physics of the ground; foundation engineering is the design that respects it\.

- **Soil mechanics**— how soil behaves under stress, built on ideas like effective stress, shear strength, and consolidation\.  Terzaghi's principle of effective stress, introduced in 1925, is the idea the whole field stands on[1](/blog/blog-geotechnical-engineering#ref-1)\.
- **Rock mechanics**— how rock masses behave, where stability depends on jointing, fracturing, and weathering as much as the rock's raw strength\.
- **Foundation engineering**— designing the footings, piles, and other elements that move a structure's load safely into the ground\.

A few more specializations sit alongside these: **slope stability** \(keeping cuts and embankments from sliding\), **ground improvement** \(strengthening weak soil before building\), and **geosynthetics** \(engineered fabrics that reinforce or drain soil\)\.  Each is a field of its own\.  For most projects, the three core domains carry the weight\.

These domains are only as good as the data behind them— which is where testing comes in\.

## The Tests: In\-Situ and Laboratory Soil Testing

Geotechnical data comes from two test families: in\-situ tests performed in the ground— chiefly the Standard Penetration Test \(SPT\) and Cone Penetration Test \(CPT\)— and laboratory tests run on recovered samples, including Atterberg limits, triaxial and direct shear, Proctor compaction, consolidation, and sieve analysis[1](/blog/blog-geotechnical-engineering#ref-1)\.  The SPT and CPT are the workhorses of site investigation— they tell the engineer how dense, strong, and stable the ground really is\.

In\-situ tests measure soil where it sits, under real conditions and real water pressure\.  Lab tests trade that realism for control, isolating one property at a time with precision\.  Good investigations use both\.

```html-table
<table><thead><tr><th>In-situ test</th><th>What it measures</th></tr></thead><tbody><tr><td>Standard Penetration Test (SPT)</td><td>Soil density and strength via blow counts</td></tr><tr><td>Cone Penetration Test (CPT)</td><td>Continuous resistance and soil layering</td></tr><tr><td>Shear vane</td><td>Undrained shear strength of soft clay</td></tr><tr><td>Geophysical methods</td><td>Subsurface profiling without drilling</td></tr></tbody></table>
```

```html-table
<table><thead><tr><th>Laboratory test</th><th>What it measures</th></tr></thead><tbody><tr><td>Atterberg limits</td><td>Soil plasticity and consistency</td></tr><tr><td>Triaxial / direct shear</td><td>Shear strength under controlled stress</td></tr><tr><td>Proctor compaction</td><td>Optimal moisture for maximum density</td></tr><tr><td>Consolidation</td><td>Settlement behavior over time</td></tr><tr><td>Sieve analysis</td><td>Particle-size distribution</td></tr></tbody></table>
```

The point isn't to run every test\.  It's to pick the ones that answer the questions this site is asking\.  That data feeds the design of two things above all: foundations and earth\-support structures\.

## Foundations and Earth\-Support Structures

Geotechnical engineering governs two big design categories: foundations— shallow \(footings, slabs\) when competent soil is near the surface, deep \(piles, caissons\) when load must reach stronger strata below— and earth\-support structures like retaining walls, soil nailing, MSE walls, and tiebacks[1](/blog/blog-geotechnical-engineering#ref-1)\.  Shallow foundations spread the load near the surface; deep foundations carry it down to ground that can take it\.

The choice between shallow and deep comes down to what's under the building and how much it will settle\.  Bearing capacity and settlement drive the decision every time\.

```html-table
<table><thead><tr><th>Foundation type</th><th>When it's used</th><th>Examples</th><th>Key concern</th></tr></thead><tbody><tr><td>Shallow</td><td>Competent soil near the surface</td><td>Spread footings, mat/raft slabs</td><td>Bearing capacity, settlement</td></tr><tr><td>Deep</td><td>Surface soils too weak; load must reach lower strata</td><td>Driven piles, drilled shafts, caissons</td><td>Load transfer, depth to capable ground</td></tr></tbody></table>
```

Earth\-support structures hold back soil where the ground can't stand on its own\.  The common types include:

- Retaining walls \(cantilever and gravity\)
- Mechanically stabilized earth \(MSE\) walls
- Soil nailing and tiebacks for cuts and excavations
- Slurry walls and gabions
- Reinforced embankments

All of it is governed by three forces the engineer keeps in balance: bearing capacity \(what the ground can hold\), settlement \(how much it moves\), and lateral earth pressure \(how hard the soil pushes sideways\)\.  Get those numbers right and the structure stands\.  All of this design intent lands in a single document— the geotechnical report— and knowing what it controls is where project owners get value\.

## The Geotechnical Report— and When Your Project Needs One

A geotechnical report tells the design team what the ground will allow: it sets foundation type, bearing limits, settlement expectations, slab support, lateral earth pressures, earthwork criteria, and drainage requirements[5](/blog/blog-geotechnical-engineering#ref-5)\.  Skip the geotechnical investigation and you don't save the cost— you defer it to change orders, differential settlement, or worse\.

In practical terms, the report is the rulebook every other discipline designs against\.  Here's what it controls:

- **Foundation type**— shallow or deep, and why
- **Bearing limits**— how much load the soil can carry
- **Settlement expectations**— how much movement to design for
- **Slab support**— how floors and slabs\-on\-grade are bedded
- **Lateral earth pressures**— loads on walls below grade
- **Earthwork criteria**— compaction, fill, and cut specifications
- **Drainage requirements**— how water is kept away from foundations

So when does a project actually need one?  Almost any new structure benefits, but it becomes essential with problem soils \(soft clay, uncontrolled fill, expansive soil\), sloped sites, a high water table, or seismic zones\.  If any of those describe your site, a geotechnical engineer isn't optional\.

The honest version: the report is one of the cheapest forms of risk insurance on a project\.  A few boreholes up front cost a fraction of underpinning a settling foundation later\.  The report has looked broadly the same for decades\.  What's changing is the data and tools behind it\.

## The Technology Frontier: AI in Geotechnical Engineering

Machine learning, digital twins, and satellite\-based monitoring are genuinely extending what geotechnical engineers can predict and observe— but operational adoption still lags the rest of the AEC sector, and for defensible reasons[3](/blog/blog-geotechnical-engineering#ref-3)\.  AI in geotechnics works best inside the data it was trained on— and the ground is rarely so cooperative\.

In research, machine learning is already being applied to real geotechnical problems: soil classification, predicting the load\-bearing capacity of shallow and deep foundations, slope\-failure early warning, and liquefaction analysis[3](/blog/blog-geotechnical-engineering#ref-3)\.  The methods will be familiar to anyone who has looked at [how AI models learn from data](/blog/what-is-generative-ai)— systems trained on past examples to predict new outcomes\.

So why hasn't it taken over?  Three reasons, and they're worth stating plainly\.

> **Why adoption lags:** 1\. **Site\-specificity**— models trained on one site's data rarely generalize to another, because the ground changes from block to block\. 2\. **Opacity**— many models are "black boxes" that can't show their reasoning, which is a hard sell for safety\-critical work\. 3\. **Extrapolation limits**— the models perform well inside their calibration data and poorly outside it, and real sites constantly serve up the unfamiliar[3](/blog/blog-geotechnical-engineering#ref-3)\.

And this isn't unique to geotechnics\.  The broader AEC sector has been comparatively slow to adopt AI[4](/blog/blog-geotechnical-engineering#ref-4), and survey estimates of how many firms actually use it in operations range widely— roughly a quarter to a half, depending on how "AI use" is even defined\.  The honest answer is a range, not a headline\.

Meanwhile, monitoring technology is moving faster than prediction\.  InSAR satellite radar can track millimeter\-scale ground movement across whole regions without a single sensor in the field, supporting dam, slope, and settlement monitoring[7](/blog/blog-geotechnical-engineering#ref-7)\.  IoT and fiber\-optic sensors feed continuous data, and digital twins turn that data into living models of real sites\.

Here's the throughline, and it's the part that matters\.  The firms pulling ahead aren't replacing engineering judgment with algorithms; they're pairing the two\.  Data tools amplify a seasoned engineer's judgment— they don't substitute for it, especially where site data is sparse\.  Adoption is [the slow, human work of AI adoption](/blog/building-ai-culture) that change always demands, and that's where most of the value gets won or lost\.  None of this changes who does the work— so what does it take to become the engineer holding the judgment?

## Careers: How to Become a Geotechnical Engineer, Salary, and Outlook

Becoming a geotechnical engineer means earning an ABET\-accredited civil or geotechnical engineering degree, passing the Fundamentals of Engineering \(FE\) exam, working roughly four years under a licensed Professional Engineer \(PE\), then passing the PE exam— a master's degree is often preferred[6](/blog/blog-geotechnical-engineering#ref-6)\.  There is no separate federal salary figure for geotechnical engineers— they're counted under civil engineers, who earned a median of $99,590 in May 2024[2](/blog/blog-geotechnical-engineering#ref-2)\.

The licensure path is well\-defined:

1. Earn an ABET\-accredited degree in civil or geotechnical engineering\.
2. Pass the Fundamentals of Engineering \(FE\) exam\.
3. Work about four years under a licensed PE\.
4. Pass the PE exam to become licensed yourself\.

On pay and outlook, the data has to be read carefully\.  The U\.S\. Bureau of Labor Statistics doesn't track geotechnical engineers as a separate occupation— the discipline is reported under civil engineers, whose median wage was $99,590 in May 2024[2](/blog/blog-geotechnical-engineering#ref-2)\.  Any "geotechnical engineer salary" you see online traces back to that civil\-engineer figure\.  Employment of civil engineers is projected to grow 5 percent from 2024 to 2034, faster than average, with about 23,600 openings each year over the decade[2](/blog/blog-geotechnical-engineering#ref-2)\.

A few questions come up again and again— here are the direct answers\.

## Frequently Asked Questions \(FAQ\)

### What does a geotechnical engineer do?

A geotechnical engineer investigates soil, rock, and groundwater conditions and designs the foundations and earthworks that let structures stand safely[1](/blog/blog-geotechnical-engineering#ref-1)\.  The work runs from site investigation through testing to design recommendations captured in a geotechnical report\.  In short, they make sure the ground can carry whatever gets built on it\.

### Is geotechnical engineering part of civil engineering?

Yes\.  It's a specialized branch of civil engineering focused on earth materials rather than the structures themselves[1](/blog/blog-geotechnical-engineering#ref-1)\.  A geotechnical engineer studies the ground; a structural engineer designs the building that sits on it\.

### What's the difference between shallow and deep foundations?

Shallow foundations— footings and slabs— spread load near the surface when competent soil is close by\.  Deep foundations— piles and caissons— carry load down to stronger strata when surface soils can't support it[1](/blog/blog-geotechnical-engineering#ref-1)\.  Bearing capacity and settlement drive the choice\.

### What is in a geotechnical report?

A geotechnical report specifies foundation type, bearing limits, settlement expectations, slab support, lateral earth pressures, earthwork criteria, and drainage requirements[5](/blog/blog-geotechnical-engineering#ref-5)\.  It's the document every other discipline designs against\.

### Will AI replace geotechnical engineers?

No\.  AI assists with prediction tasks like soil classification and slope\-failure warning, but site\-specific data and engineering judgment remain essential, and operational adoption is still early[3](/blog/blog-geotechnical-engineering#ref-3)\.  The pattern across geotechnics is pairing AI with expert judgment, not swapping one for the other\.

## The Ground Doesn't Negotiate

Geotechnical engineering is the quiet discipline that decides whether everything above it stays standing— and the firms doing it best are pairing seasoned judgment with new data tools, not choosing one over the other\.  The ground doesn't negotiate\.  The job is to understand it before you build on it\.

For a field that rarely makes headlines, the stakes are absolute: foundations, slopes, and retaining walls all trace back to whether someone read the ground correctly\.  The technology layer— machine learning, digital twins, InSAR monitoring— is real and growing, but it works by amplifying experienced engineering judgment rather than replacing it\.  That's the throughline worth holding onto\.

If your firm is weighing where AI and data tools actually fit alongside experienced engineering judgment, that's the kind of decision [an implementation partner can help you map](/services/ai-implementation)\.

## References

1. Wikipedia, "Geotechnical engineering" \(2026\) — [https://en\.wikipedia\.org/wiki/Geotechnical\_engineering](https://en.wikipedia.org/wiki/Geotechnical_engineering)
2. U\.S\. Bureau of Labor Statistics, "Civil Engineers: Occupational Outlook Handbook" \(May 2024 data\) — [https://www\.bls\.gov/ooh/architecture\-and\-engineering/civil\-engineers\.htm](https://www.bls.gov/ooh/architecture-and-engineering/civil-engineers.htm)
3. Shahin, M\., "Progression of artificial intelligence/machine learning in geotechnical engineering," Machine Learning and Data Science in Geotechnics, Emerald Publishing \(December 2024\) — [https://www\.emerald\.com/mlag/article/1/1/5/1267538/Progression\-of\-artificial\-intelligence\-machine](https://www.emerald.com/mlag/article/1/1/5/1267538/Progression-of-artificial-intelligence-machine)
4. American Society of Civil Engineers, "Architecture, engineering, construction sector slow to adopt AI, survey shows" \(2025\) — [https://www\.asce\.org/publications\-and\-news/civil\-engineering\-source/article/2025/12/18/architecture\-engineering\-construction\-sector\-slow\-to\-adapt\-ai\-survey\-shows](https://www.asce.org/publications-and-news/civil-engineering-source/article/2025/12/18/architecture-engineering-construction-sector-slow-to-adapt-ai-survey-shows)
5. turn2engineering, "Geotechnical Report: Essential for Construction Success" \(2025\) — [https://turn2engineering\.com/civil\-engineering/geotechnical\-engineering/geotechnical\-report](https://turn2engineering.com/civil-engineering/geotechnical-engineering/geotechnical-report)
6. American Society of Civil Engineers, "Geotechnical Engineering" \(certification/career, accessed 2026\) — [https://www\.asce\.org/career\-growth/civil\-engineering\-certification/geotechnical\-engineering](https://www.asce.org/career-growth/civil-engineering-certification/geotechnical-engineering)
7. Encardio Rite, "Emerging Technologies in Geotechnical Instrumentation 2018–2025" \(2025\) — [https://www\.encardio\.com/blog/emerging\-technologies\-geotechnical\-instrumentation\-2018\-2025\-part1](https://www.encardio.com/blog/emerging-technologies-geotechnical-instrumentation-2018-2025-part1)


---

Source: https://dancumberlandlabs.com/blog/geotechnical-engineering/
