Part 1: Intelligent Transit Planning & Advanced Transportation Analytics

How smart cities are building interconnected transit systems from big data and machine learning.

How smart cities are building interconnected transit systems from big data and machine learning.


As our cities become more complex, there is an increasing need to understand them better and proactively design transportation systems that meet the demands of the future. The use of analytics to support the evolution of transit is an ongoing discussion and is becoming more prevalent given the increasing amount of data that we are generating on a daily basis.

Think of your engagement with connected devices, which generate precise, real-time location data. Your smartphone. Your car’s navigation system. Sensors that track the movement of city public transport vehicles, such as buses, streetcars and subways; and pedestrian traffic along key city intersections and streets. Transit cards every time you tap on to access transit systems. All these examples yield extensive amounts of valuable data for transit planning and network management.

The amount of data provides an opportunity for both public agencies as well as private companies to improve a city’s transit system. This blog highlights many of these opportunities in five pillars: Multimodal Journey Mapping, Demand-Responsive Systems, Economically Efficient Networks, Descriptive and Predictive Insights, and Connect Data Across Organizations.  These pillars are interdependent and future smart cities will have to develop capabilities in all areas. In part 1, we focus on the first two pillars and provide you with some ideas in solving the issues within our transit system.


Multimodal Journey Mapping

Traditional transit systems, both public and private, have been planned in isolation.  Uber builds their network independently of public transportation. The MTA in New York City only oversees rail and bus travel with limited interaction with the NYC Taxi and Limousine Commission. Toronto Island Ferry operates a separate bike rental company from Bike Share Toronto. This planning does not meet today’s transportation needs mainly because people are seeking immediate access to transit systems to save time and make their trips more efficient.  

Modern transit systems understand that people are taking trips across transportation modes that vary depending on the destination, time and weather conditions. Cities must use modern analytic techniques to bring their siloed systems together to enable the most efficient trip planning.

Transit planners today must consider people’s multimodal trips, which frequently involve systems from multiple agencies, companies as well as personal vehicles.  

A typical commute for a Toronto worker could include driving their car to a GO station, taking the train before transferring to a subway line and then using Toronto Bike Share for the last two kilometres of the trip.  Without end-to-end journey mapping, urban planners don’t know the true origin-destination route or total cost of a user’s trip. Additionally, they don’t know how the transportation network can be improved for the entire user journey. There are three main problems with planning transit systems in isolation.

  1. The information gap between agencies leads to poor capacity planning.  Without understanding what external modes travellers are transferring to, transit agencies cannot predict how their demand may change over time.  For example, the relocation of a bus route may eliminate demand for a previously busy Toronto Bike Share station, which the agency would not have predicted.

  2. Users receive suboptimal trip plans in terms of both cost and time.  Most tools show one system at a time (e.g. TTC or Uber or cycling.)  A user’s optimal route will be constructed of multiple modes of transportation, both public and private.  Data from Uber, GO, TTC and Toronto Bike Share would all have to be connected to plan this route.

  3. Infrastructure is built to reinforce single system travel.  Without understanding how travellers are transferring between systems, long-term developments are often designed to expand one mode of transportation in isolation.  Busy streets may be expanded for single occupancy vehicles rather than dedicated bus lanes that connect to other transit systems.

In Toronto, advanced analytics is being used to empower multimodal trip planning.  Engineers at Sidewalk Labs, an Alphabet subsidiary specializing in smart city planning, have built the Toronto Transit Explorer.

The tool uses data from 13 independent agencies to estimate and visualize travel times using multiple forms of transportation.  For example, travelers can input origins and destinations into the tool and receive routes with estimated travel times using a combination of Toronto Bike Share and the TTC.  The tool is built on R5, an open-source transportation program, as well as Toronto’s Open Data Catalogue. The Toronto Transit Explorer goes beyond older planning tools, including Google Maps, which generally provide trip plans using only one transit system at a time.


Today, transit planners can leverage granular data from travellers across transit systems.  This is enabled by GPS tracking, beacons, and transit card scanners. It is important for planners to use all these data sources and map the entire journey of travellers. Transit agencies will depend on advanced analytics to collect, join and describe these multimodal trips.

Demand-Responsive Systems

The multimodal system connectivity, overlayed with more granular, timely data widens the opportunities to discover meaningful strategies that support a continuous evolution of transportation systems. By managing transportation systems in real-time, cities can adapt resources to meet expected or unexpected changes in demand, offering flexibility to people during peak travel hours or when one-time events are taking place.

Studying how people move around the city during different times of the day, when different weather patterns occur, when one-time or recurring events (e.g. baseball or hockey games) take place or when accidents or unexpected road closures happen can help city officials deploy specific number of fleet vehicles, incentivize the flow of taxi or ride sharing alternatives into hotspots and even alter flow of traffic by switching the direction of streets in more congested areas, such as downtown. Ultimately, cities could consolidate a set of transportation and urban mobility strategies to be deployed depending on a set of predefined criteria to proactively manage individual mobility.   

Companies like Uber, Lyft and Tesla’s new robotaxi have created a new business model founded on the ubiquity of smartphones in urban areas and the need for short-journey transportation options.  Australia’s Bridj has expanded the model by creating a multi-passenger, demand responsive bus network.

These businesses depend on GPS enabled location data from both customers and drivers.  By leveraging the data generated from these businesses, private companies, as well as public agencies, have the opportunity to plan more efficient, dynamic transit systems.

The RITMO (reinventing mobility) project,  being led by Michigan University’s Seth Bonder, Collegiate Professor of Industrial and Operations Engineering is another great example of the power of a demand-responsive system. The project was designed to address the demand-responsive concept and overcome innovative constraints from rigid public transit systems. It builds on two key enablers, connected and automated vehicles, and leverages the tremendous progress in data science to design and operate a new generation of on-demand urban transit systems.

This concept looks to improve the extreme scenarios of high and low-density transit demand within the Michigan campus, increasing bus frequency along popular routes during peak hours and introducing smaller shuttles for passengers trying to reach less traveled, peripheral routes, all through an integrated transit system and managed through a synchronized digital platform.

The project argues that such a concept can be scaled to the cities, bringing with it easier accessibility options for people in order to improve the quality of life across many segments of the population.  

Are you based in Toronto and working on Internet of Things - IoT initiatives?

Talk to us and share your point of view! Together we can help shape the future of our communities and contribute to Canada's leading role in Artificial Intelligence.

About the Author(s)

This article was originally written by Elie Yachoui and his Master of Management Analytics teammates at Queen's University’s Smith School of Business as a blog in their Operations and Supply Chain Analytics course taught by  Prof. Anton Ovchinnikov.

Editor: Michelle Liu

Julia McKeownComment