LocationRoad corridor in Assam, Northeast India
LengthAbout 66 km of road-side cut and fill slopes
ClientPrivate client (name withheld)
DurationAbout 9 weeks, with about 3 weeks in the field
In short: we flew drone LiDAR along the full ~66 km corridor, shot 66 seismic refraction profiles (about one per kilometre), mapped the geology in detail at 7 critical locations (2.12 km), and combined all three in one ground model. The stability analysis gave a factor of safety for each critical section in dry, monsoon-saturated and earthquake conditions, and every stretch was ranked by hazard with remedial measures and protection design inputs.
The challenge
The client needed to know which slopes along the corridor were most likely to fail, how, and what it would take to hold them, so that slope stabilisation and rockfall protection could be selected and dimensioned for the critical stretches. The corridor is a hard place to answer that question:
- Very heavy monsoon rain, of the order of 2,500–3,500 mm a year, which saturates the slopes
- Seismic Zone V under IS 1893, the highest seismic zone in India
- Dense vegetation covering the slopes, hiding the ground from photographs
- Soft alluvial soils and weathered sedimentary rock in the cut sections
- Failures mainly through rain-induced shallow slides, erosion and toe scour, and structurally controlled rock failures
- A narrow, live road: every field crew had to work around traffic
Scope of work
| Item | Quantity | What it involved |
|---|---|---|
| Mobilisation | Team of 8 with all equipment | Gurugram to Assam and back |
| Seismic refraction tomography (SRT) | 66 profiles (~1 per km) | 24-channel profiles, 3–5 m geophone spacing, 9 shots per spread, tomographic inversion and interpretation |
| Drone LiDAR survey | ~66 km, ~100 m wide corridor | RTK/PPK drone LiDAR with DGPS control; classified point cloud, DTM/DSM, contours, orthophoto, cross-sections |
| Geological mapping | 7 locations, 2.12 km | Engineering geological mapping, discontinuity survey, field strength tests, RMR, GSI, SMR, Hoek–Brown strength |
| Analysis and report | Corridor-wide | Integrated interpretation, slope stability analysis, hazard ranking, remedial measures, draft and final report |
Our approach: three data sets, one ground model
No single method can characterise a slope. We planned three independent investigations, each answering its own question, and brought them together for the analysis:
- Drone LiDAR showed what the ground surface looks like: slope height and angle, benches, cracks, scarps, drainage paths and changes at the road edge, under the vegetation.
- Seismic refraction showed what lies below: the thickness of loose cover, the depth to weathered and fresh rock, and low-velocity zones that are weak or saturated.
- Geological mapping showed how the material behaves: rock mass quality, joint orientations, possible failure mechanisms and shear strength.
- The stability analysis put the three together: geometry from LiDAR, layering from SRT and strength from geology gave a factor of safety for each critical section.
Seismic profiles were placed where they gave the most information: high cuts and embankments, visible distress, scour-prone sections beside water, and the seven geology locations. Their exact positions were fixed after reconnaissance and a quick-look LiDAR model, and every profile was surveyed by DGPS so the tomograms plot directly on the LiDAR cross-sections.
Methods and specifications
| Component | Method and key specifications |
|---|---|
| DGPS control | Static control pillars about every 7–8 km tied to Survey of India / CORS; WGS 84 / UTM; about 70 ground control points and at least 30 independent check points for the LiDAR |
| Drone LiDAR | RTK/PPK survey-grade drone LiDAR; 60–80 m above ground; 3 flight lines over the ~100 m corridor; at least 100 points per m²; vertical accuracy target of 5 cm RMSE (ASPRS 2023); multiple returns for a bare-earth DTM under vegetation |
| Seismic refraction | 24-channel seismograph; 10–14 Hz geophones at 3–5 m; 9 shots per spread with a stacked 8 kg hammer source; about 20–30 m investigation depth; tomographic inversion with RMS misfit under 5% |
| Geological mapping | Mapping at 1:500 to 1:1,000; scanline and window discontinuity surveys (IS 11315); Schmidt hammer and point-load tests (IS 8764); RMR (IS 13365), SMR (Romana), GSI (Hoek–Marinos); kinematic analysis on stereonets |
| Strength parameters | Generalised Hoek–Brown (2002) with equivalent Mohr–Coulomb c′ and φ′ over the slope stress range; Barton–Bandis for joints |
| Stability analysis | Limit equilibrium (Bishop, Spencer, Morgenstern–Price) for soil and rock-mass slopes; planar and wedge checks for rock; static, saturated and pseudo-static seismic cases (IS 1893 Zone V) |
How each stream was processed
- LiDAR: raw scans and GNSS/IMU data → PPK trajectory from the base station → point cloud and strip adjustment → check against ground control and classification → DTM, DSM and contours → sections, slope maps, 3D models and KMZ.
- Seismic refraction: SEG-2 shot records → geometry and elevations from LiDAR → first-break picking → delay-time starting model → tomographic inversion → velocity section and interpretation.
- Geology: field sheets and photographs → discontinuity database → stereonets and kinematics → RMR, SMR and GSI → Hoek–Brown c′ and φ′ → stability input parameters.
Execution and schedule
The work took about nine weeks from the work order. The field window was about three weeks, with the LiDAR, geology and seismic crews working in parallel in the post-monsoon dry season.
A typical field day ran from a 06:30 toolbox talk (safety, traffic, weather and equipment check), through field work by three crews from 07:00 to 16:30 (the LiDAR crew flew in the calm morning air; the seismic crew shot in gaps in the traffic), to data download, dual backup and quality checks in the evening. The client received a daily progress report: chainage covered, profiles done, issues and the next day's plan.
The team
| Role | People | Responsibilities |
|---|---|---|
| Project manager / senior geotechnical engineer | 1 | Planning, client liaison, QA/QC, safety, review of the stability analysis |
| Geophysicist and technician | 2 | SRT design, acquisition, field QC, inversion and interpretation |
| Senior engineering geologist and assistant | 2 | Mapping, discontinuity survey, field tests, RMR/GSI/SMR, strength |
| DGCA-certified drone pilot and co-pilot / data manager | 2 | Mission planning, LiDAR flights, data download and backup |
| DGPS surveyor | 1 | Control network, ground control, SRT and geology positioning |
| Office team, Gurugram | 5 | LiDAR processing, GIS/CAD drawings, geophysical and geotechnical analysis, report |
Quality assurance
| Activity | Checks |
|---|---|
| DGPS | Closure of the static network; repeat observations on 10% of control points; RMS under 2 cm; checks against known benchmarks |
| LiDAR | Boresight calibration; strip-overlap difference under 3 cm; independent check points (ASPRS); visual QC of classification |
| Seismic refraction | Daily line and geophone tests; reciprocal-time check within ±1 ms; repeat of noisy shots; RMS misfit under 5%; second-person review of picks |
| Geology | Compass declination correction; at least 10 point-load tests per unit; RMR/GSI/SMR assessed independently by two geologists |
| Analysis and report | Model checked by hand calculation on one section; internal peer review by a senior engineer; document control and revision log |
| Data | Chainage-based file naming; dual backup (SSD and cloud) every day; master data register |
Safety and environment
- Roadside work-zone traffic management to IRC:SP:55: cones, signboards, flagmen and reflective jackets at every seismic spread and drone take-off point
- Drone operations under the Drone Rules, 2021 with Digital Sky permissions; no flying in rain, wind over 10 m/s, poor visibility or over crowds
- Rock-face mapping only after a check for loose blocks; helmets at all times; an observer watching for rockfall; no climbing without rope access
- Hammer-source safety: gloves, goggles, ear protection and a clearance zone
- No tree felling; vegetation cleared by hand only along seismic lines; all waste taken back
What we delivered
| Component | Deliverables | Formats |
|---|---|---|
| Seismic refraction (66 profiles) | Field data and logs; location map; first-break picks; P-wave velocity tomogram with interpretation for each profile; depth to competent strata; weak-zone identification; engineering parameters; ray coverage and RMS error | SEG-2, XLSX, PDF, DWG, KMZ |
| Drone LiDAR (~66 km) | Classified point cloud; DTM and DSM; 0.5 m contours; orthophoto; slope-angle and aspect maps; longitudinal profile; cross-sections (20 m in critical stretches, 50 m elsewhere); 3D models of critical stretches; accuracy report | LAS/LAZ, GeoTIFF, DWG/DXF, SHP, OBJ, KMZ, PDF |
| Geological mapping (7 locations) | Lithology and structural maps; discontinuity database; stereonets and kinematic analysis; plans and sections; RMR, GSI and SMR; c′ and φ′; field test results; geotagged photo log | PDF, DWG, XLSX, Dips |
| DGPS control | Control station descriptions, coordinate list, network adjustment report, ground control and check-point list | PDF, XLSX, CSV |
| Report | Integrated interpretation; stability analysis (static, saturated, seismic); hazard zonation (Low / Medium / High / Critical); priority list; stretch-wise remedial measures and protection design inputs | PDF and hard copies, all data on drive |
How the results feed slope protection design
The investigation was scoped so that its outputs could be used directly to select and dimension flexible high-tensile steel protection systems:
| Design need | Why it matters | Where it came from |
|---|---|---|
| Slope geometry: inclination, height, length, surface irregularity | Mesh and anchor layout, system choice, quantities | LiDAR DTM, 0.5 m contours, slope-angle map, sections, 3D models |
| Thickness of the unstable layer and depth to competent strata | Soil nail and anchor length, bond zone | SRT velocity tomograms, cross-checked with geology |
| Shear strength (c′, φ′) and unit weight | Dimensioning and stability checks | Hoek–Brown from GSI and field strength; SRT correlations |
| Rock mass structure, joint sets, block size, failure mechanism | Mesh, cable net or barrier; anchor pattern | Discontinuity survey, stereonets, kinematics, RMR / SMR / GSI |
| Rockfall source zones and fall-line profiles | Rockfall trajectory, barrier location and energy class | Geological mapping of sources; LiDAR profiles; orthophoto |
| Groundwater and seepage | Drainage alongside the protection system | Seepage mapping; saturated stability case |
| Where to act first | Prioritising protection works over 66 km | Stretch-wise hazard ranking, factor of safety and priority list |
Why it worked
- One team, three disciplines: survey, geophysics and geology planned together and positioned on the same DGPS control, so every data set lines up.
- LiDAR first, then the ground: a quick-look LiDAR model guided where the seismic lines went, putting effort where the risk was.
- Built for the design that follows: every deliverable was chosen for the protection design it feeds, not for its own sake.
- Corridor scale, short field window: parallel crews kept about 66 km of field work to about three weeks.
Planning slope protection on a hill road, railway or project site? See our landslide and slope stability service.