AI and Shear Strength: A New Approach to Failure Prevention

In geotechnical and mining engineering, shear strength represents the material’s ability to resist deformation and failure. It defines the boundary between stability and movement, governing the behavior of slopes, excavations, and tailings dams.

According to the Mohr–Coulomb criterion, shear strength (τ\tauτ) can be expressed as:

τ=c+σ′tan(ϕ)

where c is cohesion, 𝜎 is effective normal stress, and ϕ is the angle of internal friction

These parameters are rarely constant under field conditions. Changes in pore pressure, stress redistribution, or material composition can quickly alter the available shear strength. When these changes go unnoticed, the balance between driving and resisting forces is lost—often resulting in sudden slope movements or structural failures.

From Static Measurements to Real-Time Understanding

Traditionally, engineers have relied on laboratory tests such as triaxial or direct shear experiments to define shear strength parameters. While essential for design, these methods capture only a single point in time and idealized conditions. In the field, soil and tailings are subject to continuous variation due to environmental factors, construction stages, and hydraulic influences.

Artificial Intelligence (AI) is changing how this information is interpreted. By integrating monitoring data—from instruments like piezometers, inclinometers, or radar systems—with numerical models, AI can perform continuous backanalysis of ground behavior.

This process dynamically updates the shear strength parameters (cohesion and friction angle) to reflect actual field performance rather than theoretical assumptions. Shear strength becomes a living parameter, evolving with ground conditions and offering a more realistic representation of stability over time.

Lessons from Real-World Failures

Recent slope and embankment failures around the world have demonstrated how quickly stability can deteriorate when changes in shear strength go undetected. In many cases, gradual pore pressure buildup or loss of confinement reduced effective strength long before deformation was visible.

Had AI-based backanalysis and real-time data integration been implemented, subtle shifts in effective stress or displacement could have been recognized early. Predictive alerts would have allowed engineers to intervene before the onset of failure. These events underline the importance of continuous assessment and the role of AI in translating data into foresight.

Intelligent Prevention Through Data Integration

Platform DAARWIN operationalizes this AI-driven approach to stability management. By continuously assimilating monitoring data and linking it to physics-based models, engineers can:

• Detect incipient strength reductions well before instability.

• Update stability factors dynamically as field conditions evolve.

• Reduce uncertainty by aligning models with real-time performance.

  
  

In mining and infrastructure projects, this integration represents a decisive shift—from reactive analysis after a failure to predictive, preventive control based on continuous learning from the ground.

The New Era of Stability Assessment

The combination of shear strength fundamentals and AI intelligence marks a new phase in geotechnical engineering. Rather than relying on static assumptions, engineers can now interpret and anticipate how strength evolves under changing stresses and environmental conditions.

This capability redefines stability assessment: transforming isolated measurements into continuous understanding, and unforeseen failures into preventable events.

With platforms like DAARWIN, the industry moves closer to a future where ground behavior is not just measured, but understood in real time—ensuring safer, more efficient, and more resilient designs.

Discover how DAARWIN integrates AI and real-time monitoring to continuously update shear strength and improve stability analysis:https://www.saalg.com/real-time-backanalysis

PLAXIS, OPENGROUND ,