Design better Public Transit with Planning

Algorithms for business resource optimization

Design better Public Transit with Planning

Did you know Operations Research is helping design better public transit ? Did you know It is helping improve the quality of public transit and at the same time, realize large savings ?

It helps answer the questions:

  • How many buses are needed ?
  • Where should they be parked and at what time should they be sent to service the public?
  • What territory should be covered, how many routes and where should be the bus stop and terminus?
  • Etc.

In fact, for decades now, public transit is providing interesting, complex and challenging Operations Research problems.

Operations Research can tackle such complex problems. Moreover, it is needed to optimize the use of resources, realize significant savings and produce solutions compliant with the laws, regulations and labor union contracts.

What do you mean by designing public transit service?

Designing a public transit service can be described as : given a fixed budget, public transit agencies must offer the best possible service, and there are several criteria for judging the quality of the service, the most common one are:

  • Bus frequency
  • Territory coverage
  • Travel time and transfers
  • Bus punctuality
  • Etc.

Therefore, To tackle such a complex problem, it is commonly divided it into several sub-problems that will be successively solved.

We expose in this article a possible sequence of sub-problems:

  1. Establishment of the demand
  2. Network design
  3. Timetabling

These sub-problems are commonly called: Public transit planning.

1) Demand

First, the area’s public transit demand is established. It means that a public transit agency have to know how many people are willing to travel to work, go to school, to a theater, or shopping, etc. 

In a perfect scenario, It would be necessary to know the movements of all the population residing in the territory served by the public transit agency.

This demand is usually aggregated as a matrix of origin-destination (OD), where each entry in the matrix corresponds to the number of people moving from an origin to a destination.

To establish the demand, many data collection techniques can be used, we will see some of these data collection techniques in future articles.

2) Network Design

Furthermore, with the OD matrices, the desired bus routes and frequencies per route in the system (number of buses per hour) must be determined for all hours of the day and days of the week to service and cover as many territories and users as possible.

Determining the frequency of service per route is of paramount importance for generating the timetables based on a predefined quality of service, which is the next step.

Operations research is helping design better public transit networks by minimizing resources usages and maximizing territory and population coverage.

3) Timetabling

Finally, Timetables for the public are derived from the frequencies generated on the previous step.
Timetables are created by determining the departure times of all trips to guarantee their frequency and synchronize the different routes to favor connections without having the public to wait too long.

The public timetables’ compliance to service quality guidelines, rules and regulations reflect the to the quality of service of public transit agency.

From an operations research perspective, timetabling consists of solving the problem of maximizing synchronization, which means, favor passenger transfers as much as possible. This is a critical step in the planning process since the quality of the service and next scheduling steps will depend on the timetabling.

A Hot Topic !

The public transit problems have been addressed in several studies. Many formulations characterize each of the above problems, the timetabling itself has more than one formulation: generating frequencies, departure times, passenger transfers at one or multiple bus stops, etc., some formulations are even based on passenger behavior or real-time data instead of using an established demand to optimize the interaction of the public transit services at an intermodal transport network using heuristic algorithms.

Finally, an important research area in transit system planning consider the integration of these sub-problems in one integrated problem using meta-heuristics to solve it.

Time for Scheduling

Once the timetables are created, it means that planning the public transit service is complete, Vehicle and crew scheduling can be initiated. This is called:  scheduling.