Every slope on Earth exists in a state of tension between two forces: the gravitational stress pulling material downhill and the internal strength of the soil and rock resisting that pull. When stress exceeds strength, the slope fails. A landslide occurs.

This deceptively simple concept is captured in a single metric: the Factor of Safety (FoS). It is the ratio of resisting forces (strength) to driving forces (stress). When FoS is comfortably above 1.0, the slope is stable. When it hovers near 1.0, the slope is marginal. When it drops below 1.0, failure is imminent.

At Jarbits, our TerraLux platform goes beyond conventional AI-only approaches by integrating physics-based slope stability modelling directly into its intelligence pipeline. This is what we call Physics-Informed Intelligence — and the Factor of Safety is its most powerful output.

Understanding the Factor of Safety

The Factor of Safety is defined as:

FoS = Resisting Forces (Strength) ÷ Driving Forces (Stress)

In geotechnical terms, the resisting forces are determined by the cohesion and internal friction angle of the soil or rock, while the driving forces are dominated by the gravitational component along the slope. Water plays a critical role by simultaneously reducing cohesion (weakening the soil bonds), increasing the weight of the material (adding load), and generating pore water pressure that actively pushes soil grains apart.

The FoS provides an intuitive, physics-based risk classification:

FoS Range	Stability Class	Interpretation
Greater than 1.5	High Stability	Slope is well within safe limits. Routine monitoring sufficient.
1.0 – 1.5	Moderate Risk	Slope is conditionally stable. Increased monitoring and contingency planning recommended.
Below 1.0	Critical Danger	Failure is imminent or occurring. Immediate evacuation and intervention required.

What makes the FoS so valuable for disaster management is its directness: it transforms complex geotechnical analysis into a single, actionable number that emergency decision-makers can interpret immediately.

The TRIGRS Engine: Simulating the Invisible Pressure

Rainfall is the dominant trigger of landslides across India’s vulnerable terrain. But the relationship between rain and slope failure is not straightforward — it depends on how water infiltrates the soil, how pore pressure builds over time, and how the soil’s mechanical properties degrade as saturation increases. This is where the TRIGRS (Transient Rainfall Infiltration and Grid-based Regional Slope-Stability) engine comes in.

TRIGRS is a physics-based model that simulates what happens beneath the surface during and after rainfall events. It models three interconnected processes:

• Infiltration: Rainfall penetrates the unsaturated soil surface, advancing a wetting front downward through the soil profile. The rate of infiltration depends on the soil’s hydraulic conductivity, initial moisture content, and rainfall intensity.
• Pore Pressure Rise: As water fills the pore spaces between soil grains, pore water pressure increases. This pressure acts against the normal stress holding the soil together, effectively reducing the soil’s grip on itself. The relationship is governed by Terzaghi’s effective stress principle.
• Time-Dependent Evolution: Pore pressure evolves dynamically over the duration of a storm and its aftermath. A slope that is stable after 2 hours of rain may become critical after 6 hours, or even fail 12 hours after the rain has stopped as subsurface water continues to redistribute.

TerraLux integrates TRIGRS-derived pore pressure simulations with real-time precipitation data and soil moisture proxies, creating a continuously updating picture of subsurface conditions across the entire study region.

Dynamic Material Modelling: The Rain Factor

One of the most critical and often overlooked aspects of slope stability is that soil properties are not static — they change with moisture content. A dry lateritic soil in the Western Ghats has very different mechanical properties than the same soil after three days of monsoon rain. TerraLux models this dynamic behaviour explicitly.

How Saturation Degrades Soil Strength

The platform applies transformation formulas to adjust three key soil parameters based on real-time saturation levels:

• Cohesion Reduction: As the soil saturates, the bonds between grains weaken. At full saturation, cohesion can drop by up to 30%.
• Friction Angle Reduction: The internal friction angle — which governs how well soil grains interlock and resist sliding — can be reduced by up to 20% at full saturation. Water acts as a lubricant between grains.
• Unit Weight Increase: Water-saturated soil is heavier. This increased weight adds gravitational load to the slope, increasing the driving forces in the FoS equation.

Real-World Material Behaviour

To illustrate the impact, consider how two common geological materials found in Indian landslide-prone regions respond to saturation:

Material	Base Cohesion	Base Friction	Adjusted Cohesion (Saturated)	Adjusted Friction (Saturated)
Alluvium	3,000 Pa	30°	2,100 Pa	24°
Phyllite	12,000 Pa	24°	8,400 Pa	19.2°

Key Insight Up to 30% strength loss occurs at full saturation. This is why slopes that have been stable for years can fail catastrophically during a single intense monsoon event — the soil literally loses its ability to hold itself together.

Geology is Not Uniform. Neither is Our Model.

India’s landslide-prone regions span an extraordinary range of geological settings — from the young, tectonically active Himalayan formations to the ancient, deeply weathered laterites of the Western Ghats. A single slope may traverse multiple lithological units, each with different cohesion values, friction angles, and hydraulic properties.

TerraLux addresses this heterogeneity through Finite Element Modelling (FEM), which operates across three integrated layers:

• Layer 1 — Satellite Vision (Object-Based): The top layer uses OBIA-segmented satellite imagery to identify terrain units and land cover types from above.
• Layer 2 — Finite Element Mesh: A triangulated mesh is generated that conforms to the terrain surface and subsurface lithological boundaries. Each mesh element is assigned specific material properties based on geological mapping.
• Layer 3 — The Physics Engine: The Mohr-Coulomb failure criterion is applied at each mesh element to compute the local Factor of Safety. This accounts for lithology integration (assigning material properties to zones), dynamic saturation (rain actively weakening soil strength), and the resulting FoS values that classify each zone from stable to critical.

The End-to-End Intelligence Pipeline

TerraLux’s physics-informed approach is not a standalone calculation — it is embedded within a fully automated, end-to-end intelligence pipeline:

• Input: Satellite imagery (Sentinel-1 SAR, Sentinel-2 optical), digital elevation models, real-time weather data, and geological maps are ingested automatically.
• Processing: Atmospheric correction, lithology assignment, OBIA segmentation, and feature extraction (250+ geomorphological indicators) are performed through a modular, JSON-driven pipeline orchestrator.
• Analysis: The physics-based FEM solver computes Factor of Safety values across the terrain, while hydrological modelling simulates infiltration and pore pressure dynamics.
• Output: High-resolution stability maps classified from Very Low to Very High risk, accompanied by uncertainty quantification and early warning alerts.

From Physics to Policy: Actionable Intelligence

The power of the Factor of Safety lies not just in its technical rigour but in its translatability. When TerraLux generates a stability map, every risk classification is backed by transparent, physics-based reasoning:

• A district collector can see which specific zones have FoS values approaching critical thresholds.
• An SDMA can model how a forecast 72-hour rainfall event will shift FoS values across their jurisdiction.
• Infrastructure planners can identify zones where new construction will require slope stabilisation interventions.

Combined with TerraLux’s Glass Box approach — where LIME and SHAP explainability tools identify the contributing factors behind every risk alert — decision-makers receive not just a warning, but the reasons behind it. An alert might indicate: “High Landslide Risk (FoS 0.85) — Contributing Factors: Slope Angle Critical (>45°, High Contribution), Soil Saturation High (92%, High Contribution), Vegetation Cover Low (Moderate Contribution).”

Conclusion

The Factor of Safety is where physics becomes policy, where equations become evacuation orders, and where data becomes the difference between preparedness and catastrophe. By integrating TRIGRS-based infiltration modelling, dynamic material property adjustment, finite element analysis, and real-time weather data into a single, automated pipeline, TerraLux transforms the abstract science of slope mechanics into continuously updated, location-specific, actionable intelligence.

For India’s 147 vulnerable districts and the 50 million people who live in their shadow, this is the transition from knowing that landslides will happen somewhere, sometime, to knowing exactly where the ground is weakening and precisely when to act.

About TerraLux TerraLux is Jarbits’ flagship geospatial AI platform, recognised among MeitY’s Top 30 AI Solutions and incubated at SIIC, IIT Kanpur. It is India’s first physics-informed AI platform for national-scale landslide risk assessment. Learn more at jarbits.com.