Black Friday banner
Routle.ai logo

Routle.ai Methodology

How Routle.ai builds reliable, global multimodal routing intelligence

Routle.ai integrates advanced data engineering, AI-native modelling, and InfraEconomy's transport expertise to deliver accurate, transparent, scalable multimodal routing and trade insights. Our methodology is built on three foundations: comprehensive data, holistic modelling, and AI-driven processing.

What makes us special?

Routle.ai does not depend on individual carrier offers. It uses averaged and reconstructed parameters of cost, time and infrastructure performance across countries. This gives logistics professionals a flexible tool for building and rebuilding multimodal routes without being limited to linear paths or single-provider options.

Yes, you've heard us. Routle is a proper 'helicopter view' solution for your logistics arsenal. We will not provide you with the rates from the 'usual suspects.' We believe that we fill the gap in the toolbox of every logistics professional. By deriving our methodology from macroeconomic analysis and decades of transportation geography theory, we are able to get you a key service to support high stakes decisions.

As a result, users can explore the fastest or cheapest variants, set required ports or borders, avoid certain countries and design circular or triangular routes that improve asset utilisation. Routle.ai supports this process by revealing many routing possibilities instead of just one. It helps teams discover better routes, test new corridor ideas and work with greater freedom, alternative choices and creative control over how cargo moves.

Routle.ai works similarly to familiar maps services that allow users to build passenger routes from home to work or between cities – but for cargo. Routle.ai builds routes using a global transport-network map, in which every segment contains aggregated and averaged data on transport speed and cost.

Our network graph includes:

  • 600+ thousand km of roads
  • 400 thousand km of railway lines
  • 30+ thousand maritime routes
  • 700 sea ports
  • All key border-crossing points open for cargo movements
  • The largest inland multimodal terminals

Routle's Routing In Detail

Routes are constructed between major cities and regions and do not include last-mile delivery.

For each O-D pair, Routle.ai generates several route types:

  • The cheapest
  • The fastest
  • The optimal in terms of time-and-cost trade-off
  • The optimal using only maritime transport, without inland connections (where applicable)
  • The optimal using only land transport, without a sea leg (where applicable)

A similar set of route types can be generated for containerised cargo, bulk, and liquid bulk, depending on the product. If a user changes anything in the query, e.g. avoiding countries or routing via selected transport nodes, all these routes will consider these changes.

Time-and-cost optimal routes are based on an assessment of inventory costs. The more expensive the product, the higher the contribution of delivery time to the optimal route choice. A user may enter the product value manually or leave the field blank. In that case, the system will use an average product value for the given origin and destination.

During routing, we account for a variety of constraints, for example, closed borders between certain countries or restrictions on specific cargo types. Information on softer or contextual constraints along the route can be provided by our AI-agent, Sinbad. We update our data regularly and also introduce adjustments in response to rapid changes in tariffs or new long-term transport restrictions.

Comprehensive, Multi-Source Data Foundation

Our platform combines data from a wide range of international, national, commercial and proprietary sources covering trade flows, infrastructure, operational performance, costs, speeds, technical characteristics and capacity constraints. These include:

  • Global trade statistics (Sources include UN ComTrade, Eurostat, national customs)
  • Transport and infrastructure datasets from ports, railways, highways, border operators, and investment project documentation regularly updated by InfraEconomy team
  • Geospatial and technical layers (OSM, ORM, satellite images recognition using AI-tools)
  • Operational signals from AIS and data providers for maritime infrastructure and traffic
  • Freight-rate platforms and marketplaces
  • Proprietary InfraEconomy datasets of validated infrastructure parameters and bottleneck assessments based on continuous research including field missions, interviews with market players

Data processing pipeline designed for uncompromised global coverage

Routle.ai is designed for full global coverage, including regions where publicly available data is scarce or fragmented. Limited data availability does not impede modelling: through controlled estimation techniques, cross-source validation, and AI-assisted reconstruction, the platform ensures that even data-poor geographies are represented with coherent, high-quality parameters suitable for routing, benchmarking and strategic scenario analysis.

Cleaning

  • Automated detection of outliers, duplicates, and mechanical errors
  • Correction of inconsistent units and redundant entries

Human Expert validation

  • Manual resolution of inconsistencies
  • Verification of infrastructure characteristics and capacity indicators

Transformation and modelling

  • Extrapolation and analogue-based estimation to infer missing values using comparable corridors or assets
  • Indirect inference where no direct data exists, relying on secondary indicators and statistical methods
  • Model-driven parameter reconstruction to generate consistent estimates for speeds, capacity, dwell times or costs

Data integration

  • Unified indices for infrastructure objects
  • Cross-mode linking of ports, roads, railways, borders and regions
  • Alignment with country and subnational identifiers

This results in a coherent, continuously updated global dataset that underpins all routing and modelling outputs.

Holistic Global Routing Model: From Established Corridors to Under-Served Directions

Traditional routing tools are often limited to corridors with extensive commercial data or established transport services. Routle.ai employs a holistic global transportation assignment model InfraForecast™ also created by InfraEconomy. The model is capable of evaluating any country–country or region–region pair, including directions where regular services are absent or information is limited.

Our model incorporates:

  • Multimodal networks (road, rail, maritime, inland water)
  • Country-to-country and region-to-region trade matrices
  • Capacity and bottleneck constraints for ports, borders, roads and railways
  • Speed profiles, tariffs, dwell times and typical operational delays
  • Seasonality and reliability factors
  • Scenario projections for future infrastructure or regulatory changes

This approach enables consistent evaluation of both high-density transport corridors and under-explored routes, supporting operational routing decisions and long-term strategic planning.

AI-Native Architecture Powered by "Sinbad"

Routle.ai's modelling framework is supported by Sinbad, our AI agent designed specifically for multimodal transport analytics.

Sinbad provides:

  • Automated data extraction from structured and unstructured sources
  • Classification and enrichment of infrastructure and operational parameters
  • Real-time monitoring of regulatory updates, infrastructure announcements and disruptions
  • AI-assisted completion of missing or inconsistent data
  • Dynamic refinement of corridor performance metrics, including speed, cost and reliability

Enterprise-Grade Customisation: Granular Commodities, Specialised Assets and Network Optimisation

For enterprise clients, Routle.ai offers advanced modelling capabilities tailored to specific operational contexts.

Granular commodity modelling (HS-4 level and above)

This enables:

  • Commodity-specific routing
  • Incorporation of handling requirements and restrictions
  • More precise cost and performance estimation

Specialised transport assets

Our model supports:

  • Custom truck configurations
  • Specialised containers
  • Greater variety vessel types
  • Specific technical constraints

Network-level analysis and optimisation

Beyond shipment-level routing, we provide tools for:

  • Hub and corridor selection
  • Fleet deployment strategies
  • Optimisation of multimodal transport networks
  • Stress tests and resilience assessments
  • Long-term scenario and investment planning

These capabilities help organisations refine both day-to-day operations and strategic supply-chain decisions.