Research · working paper
Re-Engineering Inland Transport Risk: A System-Powered Inspection Stack for Tail-Risk Detection in Indian Over-Dimensional Cargo Movement
Varun Gowariker, Murugan Lakshmanan and Bhavani N.
The paper in numbers
Real counts from the two repositories, read on 18 October 2026. The Zenodo count refreshes live when your browser can reach Zenodo.
| Repository | Views | Downloads |
|---|---|---|
| SSRN | 97 | 26 |
| Zenodo | 267 | 5 |
Counts change over time and each site counts views and downloads in its own way, so compare within a repository rather than across them. Zenodo's file download is restricted on the record page, which keeps its download count low.
The question
Risk on Indian roads is mostly judged from claims, insurer loss runs and retrospective surveys. They arrive after the cargo has moved, and they miss how risk builds during an over-dimensional or over-weight movement.
Method
Findings, drawn
Each card states the finding, draws it from the 2025 inspection file, and says what an inspection record changes. Two of the charts are interactive.
Tail risk (P90)
Finding. Tail delays, not averages, explain volatility.
Reading the chart. The mean delay across 242 inspections is 1.41 days and P90 is 2, yet the slowest reached 11 days. Drag the percentile to see how thin the tail is and how far it runs.
What inspection changes. An inspection record per move shows which loads, routes and carriers populate the tail, so they can be acted on before dispatch.
Concentration (HHI)
Finding. Operational concentration exceeds financial concentration.
Reading the chart. By carrier the shipment index is 2,518 and the exposure index is 2,233; both are high. Routes (408 and 422) and consignees (347 and 329) are well spread. Below, move the shares and watch the index respond.
What inspection changes. Dependence on a few carriers is greater than the money alone suggests, so governance controls belong on the carrier panel.
Frequency against exposure
Finding. High exposure does not imply high frequency.
Hover or tab to a mark for its value.
Reading the chart. Each bubble is a carrier. Carrier 03 made 18 inspections yet carries about ₹34 Cr and a P90 delay of 4.5 days; Carrier 09 made 6 and has the worst P90 at 5.5 days.
What inspection changes. Ranking by shipment count would miss them. Exposure and tail weighted inspection catches them.
Routes
Finding. Low-traffic routes can show disproportionate delays.
Hover or tab to a mark for its value.
Reading the chart. Pune to Golaghat has two inspections and a P90 of 10.7 days; Pune to Hyderabad has one at 8 days. The busiest route, Pune to Ahmedabad, holds at 2 days across 33 inspections.
What inspection changes. Thin routes get less attention and less history. A standard inspection gives them the same evidence as a busy one.
Carrier by route heatmap
Finding. Loss potential clusters in specific carrier and route combinations.
Hover or tab to a mark for its value.
Reading the chart. On Pune to Surat the P90 is 1.0 day for one carrier and 3.8 days for another. Neither the route nor the carrier explains it alone.
What inspection changes. Inspect the pair, not the parts: the same route can be safe with one carrier and exposed with another.
Consignee concentration
Finding. Consignees aggregate exposure across routes and carriers.
Hover or tab to a mark for its value.
Reading the chart. The top ten of 89 consignees hold 48% of exposure and the top 30 hold 76%.
What inspection changes. Demand-side exposure shapes contract design and inspection intensity, which claims-only models underweight.
Programme growth
Finding. Inspection-native data accumulates as the programme scales.
Hover or tab to a mark for its value.
Reading the chart. One inspection in January, 43 in September and 16 in December. Each one adds a structured record to the tail and concentration views.
What inspection changes. The more movements inspected, the sharper the picture of where the tail sits.
Computed by iCargoFinsure from the anonymised 2025 inspection file (242 records, 28 Jan 2025 to 20 Dec 2025). Identifiers are replaced by rank labels. See the paper for the authors' own figures and method. Open the full analytics.
What it means
See where exposure concentrates and which pairs of carrier and route run hot, then price and condition cover on evidence.
Spot the quiet route or rare operator that carries outsized value, before it becomes a claim.
A standard inspection becomes the data source: every record feeds the same tail and concentration views.
Fix the constrained stage once and stop delays from compounding down the chain.
Cite
Zenodo DOI 10.5281/zenodo.18044161. Keywords as listed on SSRN.
Gowariker, V., Lakshmanan, M., & Bhavani, N. (2025). Re-Engineering Inland Transport Risk: A System-Powered Inspection Stack for Tail-Risk Detection in Indian Over-Dimensional Cargo Movement. Zenodo. https://doi.org/10.5281/zenodo.18044161
@misc{gowariker2025inspection,
author = {Gowariker, Varun and Lakshmanan, Murugan and Bhavani, N.},
title = {Re-Engineering Inland Transport Risk: A System-Powered Inspection Stack for Tail-Risk Detection in Indian Over-Dimensional Cargo Movement},
year = {2025},
publisher = {Zenodo},
doi = {10.5281/zenodo.18044161},
url = {https://zenodo.org/records/18044161}
}Questions
It argues that Indian inland transport risk should be read from inspections made before and during a move, not only from claims and loss runs made after it, and shows how tail delays and concentrated exposure can be spotted earlier in over-dimensional and over-weight cargo.
On SSRN at papers.ssrn.com/abstract=6067067 and on Zenodo at zenodo.org/records/18044161 under DOI 10.5281/zenodo.18044161. The Zenodo record is licensed CC BY 4.0.
P90 is the 90th percentile: the delay that nine in ten movements beat. It isolates the slowest tenth, where most of the cost and surprise live.
The Herfindahl-Hirschman Index adds up squared shares. It runs from near zero for many equal actors to 10,000 when one actor carries everything.
They are computed by iCargoFinsure from the anonymised 2025 inspection file of 242 records. Carriers, consignees and vehicles are replaced by rank labels. The readership counts come from SSRN and Zenodo. For the authors' own figures and method, read the paper.
The platform captures the structured inspections, maker and checker review and unloading evidence that the framework relies on. The Field Academy lets you practise the same steps.
From paper to practice
Practise the maker, checker and unloading steps in the Field Academy, or book a live demo of the platform.