Dynamic optimization of intermodal city transportation

DOCIT is a multi-modal journey advisor capable of hedging risk and increasing the likelihood of arriving on time.

The system implements functions for the use of both travelers and transport network operators.

Multi-modal travel refers to combining, as part of the same trip, multiple transportation modes, such as buses, trams, and bicycles. Multi-modal travel is an important application, given its potential to fight increasing levels of congestion and pollution in urban environments.

Existing multi-modal planning systems are deterministic. This implicitly assumes that the input data, such as public transport schedules, are accurate. In reality, however, transportation networks feature many types of uncertainty, including variations in the arrival and departure times of public transport vehicles, such as buses and trams. Even small delays in a given segment of a journey could result in a missed connection, with substantial delays in the arrival at the final destination.

Journey plans computed with our system offer an increased likelihood of arriving at the destination on time. The system employs advanced algorithms capable of reasoning about the uncertainty about the network status. If a plan contains an uncertain connection, an alternative safe option is included in the plan, so that the journey can continue on a reasonably good pathway.

Functions for travelers, available through a mobile app, include multi-modal journey planning, tracking an active journey, receiving real-time updates from the network, detecting plans invalidated by events in the network, and replanning.

Functions for operators, accessible through a console available on its own or integrated within IBM IOC, include the aggregation of network KPIs from historical data, as well as a global view of current active journeys.

Besides the "standard" journey planning, which computes a plan between two given points A and B, there are many ways of using our system. For example, the What's On mobile app combines multi-modal travel with point-of-interest selection. The app can recommend a list of points of interest (e.g., "Chinese food places"), reachable with multi-modal travel within a given time (e.g., "half an hour"), from the current location of the user.


Adi Botea


Using public transport networks in a more reliable way considering:


Examples of results

Transportation screenshot request

Specifying a journey request through a mobile app.

travel distrib

A journey plan with potentially more than one alternative is provided and visualised.

travel distrib

Showing possible arrival times with their corresponding probabilities, as well as the best-case, the most-likely, and the worst-case potential trajectories.

Application areas

Smarter water
Smarter energy
Smarter healthcare
Smarter transportation
Smarter cities