Harnessing the Power of AI & Machine Learning in Geotechnical Engineering

All over the news you have heard about Chat GPT and AI. You also have heard how AI could improve productivity and increase workers output in all industries. This is true. In fact, this is an under estimation of what AI can do. Most Focus is about how Large Language Models (LLM) can help. However, LLMs like Chat GPT rely on something that is far more powerful than what you hear in the news. They rely on a specific algorithm called Neural Networks. Neural Network is a computer programming algorithm that learns data in a manner similar to how biological neurons do. In fact all what Chat GPT is a next work predictor. Its Neural Network tells it what the next word would be based on the previous ones. This is not to take away from how useful and powerful Chat GPT is. Its Neural Network is so massive in size that it requires a vast array of NVIDIA GPUs to run and train on data. Neural Network, the algorithm can help in many other ways. it can learn any function, any data, or anything in ways the human brain cannot. It can learn and make accurate predictions without the need to formulate a rigorous mathematical formulation for the observed data. Universal Approximation Theory states that neural networks with sufficient layers and neurons can learn any continuous function at any desired accuracy. This makes Neural Networks a very powerful regression tool.

In this article, the uses Neural Networks and Large Language Models is illustrated. Useful applications are discussed in detail so that geotechnical engineers can apply to their workflow to ensure they can get the most of this powerful technology. Other people of different specializations can also learn from this article and apply to their workflow. What is illustrated here are already applied in Geo-Mechanica in consultations, software development, and day to day tasks.

Large Language Models

Most people probably have used Chat GPT and asked it different questions. in work, people probably

would ask it to write, review, and summarize emails, articles, reports, or presentations. when they get home, people probably would ask the same technology how many calories in my food, how to improve their lifestyle, what is the recipe for making a delicious apple pie. Financial Analysts would go as far as using it to analyze stocks, financial statements, and market news summaries. Scientists would ask it for best regression model to use, what can be inferred from data, and how to improve experiments. Now Most likely, the reader here is a Geotechnical Engineer. So Geotechnical Engineers can ask it to summarize soil reports, suggest a foundation design, and suggest a soil improvement program. these are some of the things A user of LLMs would ask and there are many others.

The problem with general LLMs like Chat GPT is that they are trained on all the internet. The Language is general; the information can be corrupted and inaccurate. There is no guarantee that the prompt would yield a correct output. This in addition to hallucinations can make LLMs output less reliable in many cases. Hallucinations is the tendency of LLMs to give very weird and very wrong output when the context is very long for them. give Chat GPT a one line question it can give a very good answer. give it a 100 pages report, you might be surprised by how many errors in the summary it gave you. You might even find information that was not in the report to begin with. In fact any LLM user should always review the output of an LLM before making any decision based on the output.

This is where Retrieval Augmented Generation (RAG) LLMs are superior to general LLMs. RAG models are a type of LLMs that would give output based on the data they are trained on and given to them. When asked to give information outside their training data, they don’t give an answer. This reduces the chances of hallucinations significantly. One such tool is called Notebook LM developed by google. In Notebook LM, you can give it many types of documents such as websites, PDFs, or even a YouTube Video. You can then ask it to summarize, review and suggest improvements based only on the input you have given it.

In Geo-Mechanica we created a RAG LLM called Dan trained on many geotechnical design codes and books. it has been very useful for Geo-Mechanica in retrieving information from these codes which is far more faster and efficient than scrolling PDFs or flipping pages. In addition, Dan can do many things for Geotechnical Engineers such as suggesting designs, foundation layout, and soil profile analyses. A free version is made available free here on the website (lower right corner). Go ahead and ask it and tell us what did it tell you. Not only geotechnical help does dan provide, but office work such as searching internal documents, reports, and data. The more experience we gain, the more experience it gets. You can also contact us for access to the full powerful version. we can also develop custom Dans tailored to you company and needs.

Let’s see an example of using Dan. It was given an image showing a real soil profile. It was then told that the soil will be excavated and supported. The following prompt was used:

What is the best excavation support method based on the soil profile given is there any concerns I should look closely into. and what is the typical number of struts I should use in this case.

It gave the following reply:

Based on the soil profile shown in your diagram, I can provide recommendations for excavation support methods and key considerations.

Recommended Excavation Support Method

For your soil profile with alternating clay (CL) and silty sand (SM) layers and an excavation depth of approximately 12.3m, I recommend a braced excavation method with the following components:

Primary Support System:

·       Diaphragm wall or secant pile wall for the retaining structure

·       Horizontal struts with wales for internal bracing

·       Center posts to support the strut system

Alternative Options:

·       Soldier pile and lagging with horizontal struts (if groundwater is not a major concern)

·       Sheet pile walls with bracing (though less suitable for this depth and mixed soil conditions)

Number of Struts Required

For a 12.3m deep excavation, you should typically plan for 3-4 levels of struts:

Strut spacing = 3-4 meters

Typical arrangement:

·       1st level: 2-3m below ground surface

·       2nd level: 5-6m depth

·       3rd level: 8-9m depth

·       4th level: 11-12m depth (if needed)

The exact number depends on:

·       Strut load capacity

·       Wall stiffness

·       Soil strength parameters

·       Allowable wall deflections

Critical Concerns to Address

1. Mixed Soil Conditions

The alternating CL and SM layers create several challenges:

·       Differential settlements between layers

·       Varying earth pressures at different depths

·       Potential for piping at sand layer interfaces

2. Groundwater Management

This is likely your most critical concern:

·       Sandy layers (SM) may have high permeability

·       Dewatering will be essential to prevent:

o   Sand boilingdrained shear strength of clay

·       $$H$$ = excavation depth

·       $$q_s$$ = surcharge loads

4. Wall Penetration Depth

The wall should penetrate sufficiently below excavation level:

·       Minimum 0.7H to 1.0H below excavation base

·       Check both push-in stability and basal heave

Design Recommendations

Earth Pressure Distribution

Use apparent earth pressure diagrams for strut design:

·       For mixed conditions, consider the more conservative clay pressure distribution

·       Apply appropriate factors of safety (typically 1.5-2.0)

Construction Sequence

1.      Install retaining wall first

2.      Excavate in stages (2-3m lifts maximum)

3.      Install struts immediately after each excavation stage

4.      Implement dewatering system before excavation begins

Monitoring Requirements

·       Inclinometers in the retaining wall

·       Settlement monitoring of adjacent structures

·       Piezometers for groundwater monitoring

·       Strut load monitoring

The mixed soil profile requires careful attention to groundwater control and staged construction to ensure stability throughout the excavation process.

You can try it yourself. It is very smart. When asked about how many struts and how thick a diaphragm wall should be is used, it replied with 3 to 4 struts and the wall should be 0.6 m thick. this was a very good estimate compared to the case study given to it.

So, to wrap it up what LLMs can do, it could be said that for the first time experience and knowledge be it public or internal to a company can be retained and reasoned with in a very human way to help shape up future decision making. Dan will become more experienced, the more experienced we get. It will become more knowledgeable into the ways of Geo-Mechanica as a company. It will also become more knowledgeable in Geotechnical Engineering as more and more reports, analyses and data are fed into its Neural Network.

There are multiple levels of using LLMs for Geotechnical Engineer beginning from day-to-day work (Level 1) to providing insights and suggestions on designs and decisions (Level 4).

Here are some Level 1 cases for Dan and RAG LLMs in general:

• Reply to emails

• Review reports for linguistic errors

• Summarize reports, emails and give you suggestions for next steps based on those summaries

• Schedule your work based on importance and deadlines

  
  

Level 2 is more specialized and here are some Level 2:

• Show it a soil report or a geologic model and ask what some interesting parts to look for or what hazards are there.

• Ask it to bring you case studies from literature similar to a project you are working on.

• Ask it to provide you with relevant local code guidelines for a specific soil profile, retaining wall, or a foundation for a building.

  
  

In Level 3 Dan can get more involved:

• Ask it to review a design you came up with and check if it meets relevant codes or is it suitable for the soil.

• Ask it to review your calculations and if there any errors or inconsistencies here or there.

• Ask it the calculations method used here is suitable for the problem and if there are better methods to make these calculations.

• Ask if the soil profile adopted is good and does not ignore important features.

In Level 4 the Ai can do more involved work, make calculations itself, and give a report that you can then review and approved. No decision should be left out for AI to make with human insights as these models can make errors. However, to improve the output of the model, previous calculations and reports can be given to it so that it can have a solid basis to begin with. Large tasks should be divided. for example, don does not ask Dan to make a full report that is 40 to 50 pages. Instead, make subtasks for it to work on. Task 1 could be soil profiling. Task 2 would be giving it the same soil profile it made for foundation suggestions or excavation support. Task 3 bearing capacity and settlement calculations for foundations or wall and ground deformaiton for excavations.

Neural Networks

Artificial Neural Networks is the algorithm behind LLMs, but their applications exceeds far more than that. They can be very powerful in creating models and give predictions in many areas of Sciences and Engineering. In Geotechnical Engineering, A Neural Network can be trained to predict lateral wall deformation based on case studies or a firm's own measured data. the Neural Network can then be used to design future excavation support. This is just one example. Decision making Processes can also be quantified and put into a Neural Network to facilitate future decision making. Many more applications can benefit from the powerful algorithm.

In Geo-Mechanica We built ML Lab which a Neural Network building software that can be used to train neural networks on all sorts of data. It is a free tool and is available for site subscribers here. A step-by-step video tutorial of how to use ML Lab is provided in this video. Try it. It is very easy to use. You are literally 3-6 clicks away from building your first neural network.

Video shows a step-by-step tutorial on how to use ML Lab for building artifical Neural Networkss

Here are some suggestions of how Neural Networks can be utilized as an analysis and design tool for geotechnical engineers:

• Train Neural Networks on different foundation designs you have come up with from your previous work. This is very useful when using piles and piled rafts as the number of variables involved is very large. Use the trained model on new designs to come up with design alternatives much faster.

• Similar to the example in the video shown above, you can train a Neural Network on case studies of excavations.

• Your corporate has made decision that has been made that led to desired outcomes and some have led to undesirable outcomes. quantify the input that led to those decision. Train a neural network that can help future decision makers in your firm make informed and better decisions. After so many decisions the neural network can be very smart that decision making can be automated and just reviewed by the seniors.

• If you are working with Finite Element Codes and run analyses, you know how time consuming it can get sometimes to model complex problems. Train a Neural Network on results of previous Finite Element analyses so that it can inform on future model building.

Step by Step Example on Training a Neural Network on ML Lab

ML Lab is free and can be downloaded from here. Once downloaded, install it and run the program. In this example, a Neural Network is trained on data of measurement of diaphragm wall maximum deformation as a result of excavation. The variables that were used to inform the network are 13 variables (e.g. excavation depth, number of struts, ...etc. ). The output of the neural network is the maximum wall deformation as shown in Fig. 1 amd a sample of data points are shown in Table 1. You can also watch the example from the video here.

Figure 1 showing variables used in training the Neural Network and the output which is the maximum lateral wall deformation (red).

Table 1 showing samples of data points from excavation data

Construction method	Total excavation stages	Excavation width (m)	Final excavation depth (m)	Wall length (m)	Wall thickness (m)	EI (MN-m²/m)	No. of struts	Avg spacing	Depth of hard stratum (m)	ΣΗc/Hwall	Ei / σ'	Su / σ'	δₕₘ/Hₑ (%)
1	7	33.4	18.5	31	0.8	918	6	3.08	31	0.87	650	0.1	0.33
1	4	29.3	10	18	0.6	447	3	3.33	42	0	0	0	0.22
1	6	33.4	16	27	0.7	709	5	3.20	55	0	0	0	0.52
1	5	49.3	12.3	21.5	0.6	447	4	3.08	46	0	0	0	0.37
1	5	35	13.2	28	0.8	1059	4	3.30	45	0	0	0	0.44
1	5	54.1	12.6	22	0.6	447	4	3.15	48	0.75	650	0.31	0.48
2	7	41.2	19.7	35	0.9	1507	7	2.81	46	0.87	650	0.31	0.54
1	7	35.5	18.1	40	1	2067	6	3.02	46	0	0	0	0.42

Couple of things are needed before plugging in those data into ML Lab. firt you need to make sure to normalize the output values by dividing over the maximum value to make the output falls between zero and 1. This is because the neural network is better trained when the predicted values are not larger than 1. There are things that should be done for other issues that may arise in data and are further discussed in the manual and in video 1. Next thing is that data must be in CSV format. If you are using a spreadsheet, you can manipulate your data as you wish but final data should be exported to a CSV file.

Now open ML Lab and import your data file, insert the number of input columns and the output columns (only one output is allowed) as shown in Fig. 2.

Figure 2 showing ML Lab interface and how to import and adjust neural network settings.

In the Neural Network hyperparameter tab page (Fig. 3) you can adjust the number of layers, the number of neurons per layer, the learning rate, the epochs, and many more. Once you adjust those click on the Neural Network button just below the file path entry and wait for training to finish.

Figure 3 showing neural network hyperparameter tuning page.

If you want to use the neural network, you will find the file path at which it was saved in the log. copy the path, check the use the neural network for prediction checkbox and fill in the path to the prediction file and the neural network.

Conclusion

Machine Learning and AI are transforming how everything is being done in all specialties. They can increase productivity and minimize time needed for many tasks. LLMs and RAG LLMs use cases have been presented and categorized based on depth on involvement of the LLM.

Most people think of Chat GPT as the only thing related to AI. However, AI and machine learning are much more than that. Chat GPT is an LLM that uses Neural Networks to do the magic it is doing. Neural Network is a powerful algorithm that can be used in many ways. Examples of use cases have been given in this article and an example of using Geo-Mechanica ML Lab software to train a Neural Network was given as a video tutorial.

Disclaimer:

AI/ML outputs (including LLMs/RAG/neural networks) are provided for informational purposes only and are not professional engineering or legal advice. Geotechnical decisions must be based on project-specific data and independently verified using accepted methods and applicable codes. Any designs, reports, or recommendations must be reviewed and, where required, sealed by a qualified licensed geotechnical engineer. Do not input confidential or proprietary project information into AI tools unless authorized and compliant with your policies and contracts. The authors disclaim liability for outcomes resulting from reliance on AI-generated content. “AI outputs must be checked against local codes, project constraints, and professional judgment.