Toronto Cycling App for the Cycling Network Plan Jacquelyn Hayward Gulati Manager, Cycling Infrastructure and Programs Big Transportation Data for Big Cities – June 14, 2016

www.toronto.ca/cycling - City of Toronto Cycling www.toronto.ca/cyclingapp - Cycling App www.toronto.ca/cyclingnetwork - Cycling Network Plan @TO_Cycling

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Toronto Cycling App City of Toronto contracted developer Brisk Synergies to build upon their Open Source CycleTracks app developed for the city of San Francisco 

Effective and inexpensive method of collecting data



Gathers anonymous GPS trip data and demographic information to enrich data



Public engagement and consultation tool to gather cyclist input in order to plan cycling routes



2014 - 55,000+ bicycle trips recorded using the App, 3,200 App users

Cycle Tracks being used, either native or rebranded: Austin, TX, Atlanta, GA, Montreal, QC, Reno, NV, Philadelphia, PA, Monterey, CA, Raleigh, NC, Fort Collins, CO, Minneapolis/St Paul, MN, Seattle, WA, Salt Lake City, UT, Los Angeles, CA, Lexington, KY

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Toronto Cycling App: Features 

GPS Route Mapping



Real-time Statistics: average, current, max speed and distance during trip



Cumulative statistics by trip type



Calories Burned, Greenhouse Gas Offset

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Toronto Cycling App: Map Features



Road Restrictions / Construction



Water Fountains



Traffic Data



Bicycle shops



Real-Time BikeShare data

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Innovation Impact & Transferability Innovation: 

First time the City has used GPS technology to engage citizens in cycling network planning



Very strong uptake from citizens

Community: 

Cycling community has been engaged, anxious to see the results and supportive of the approach



Data was made available as part of overnight “TrafficJam Hackathon” in partnership with Evergreen CityWorks (Fall 2015)

Transferability: 

App is to be open source code, with customization



Other municipalities can implement a similar project

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Data collection methods considered Options Considered 

Balance between cost / data precision and quality



Strava?

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Privacy, User survey & trip data collection

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Road Centreline network vs Cyclist Routes Cyclists do not follow the road network 

GPS traces across parks, parking lots, driveways, campuses, and other places not coded in the City of Toronto's street centreline network GIS file.



GPS traces not aligning exactly with the network links.

Process 

Each route of travel manually input in to road centreline shapefile to prepare it for the GPS mapmatching process



Success of the map-matching algorithm is dependent on the completeness of the road network shapefile

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Data Cleaning & Preparation 1.

Clean the data (e.g., remove outlying signals, signal errors, or very short segments)

2.

Create a complete bicycling network that includes the network of streets and other links that cyclists use to travel - which may not be included in road shapefile (e.g., park trails, parking lots, and driveways),

3.

Match the GPS points collected for each bike trip to the correct network links.

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Combating Data Bias App data is inherently biased towards people who have/can afford a smartphone 

Bias vs quantity



Bias can be mitigated if sample size is large enough



Which biases are acceptable? (does income level affect cyclist route choice?)



What biases can we undo with technology (upload over Wifi vs mobile data plan. iPod vs mobile phone)



Incentives: try to increase representative sample and encourage larger demographics of people to participate



Encourage representation from broad spectrum of cyclists from varying income, gender and cycling experience levels



Increase suburban cyclist participation: Scarborough, Etobicoke, North York

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

2 years : 4,000 Users 90,000+ Trips Being Analyzed

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Current Cycling Network

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

City of Toronto Cycling Network

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

City of Toronto Cycling Network 

54% of Toronto adults are cyclists



35,000 daily bike trips into core



700,000 bike share trips per year



4,000 cyclists / day on College and Harbord bike lanes

Cycling is growing in all districts.

** statistics from prior counts/surveys

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Connect – Grow – Renew: Building Toronto’s cycling network over the next 10 years. 10yr Cycling Network Plan 

Connect the gaps in our existing Cycling Network;



Grow the Cycling Network into new parts of the City; and



Renew the existing Cycling Network routes, where there are opportunities to improve their quality.

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

App Data Analysis – Key Findings Assumptions were validated Desire Lines, direct routes 

Cyclists want & take direct routes

Impact of Cycling Infrastructure 

38% recorded trips within the limits of the current network



10% of routes have existing cycling infrastructure

Network Usage – Mean Volume of Trips per route segment 

37.4 trips/route = no infrastructure vs 207.7 trips/route = Infrastructure

Arterials vs Quiet Streets 

31% Local Streets, 54% Arterials



Cyclists more likely to use signed local roads than unmarked roads

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Data Analysis Barriers 

Routes crossing rail corridors, highways, rivers – twice mean traffic volume than average (110 vs 54)



Reinforce need to provide good barrier crossings.

Trail usage Scarborough/Etobicoke/North York – trails are an important part of the network backbone 

13% of all trips were made on trails



16% on street routes

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Data Analysis -> Network Planning Most Heavily Travelled Routes with no Cycling Infrastructure: 

Bloor: Dundas St West – Sherbourne



Danforth: Broadview – Woodbine



College Street: Dundas – Manning



Queen Street W: Roncesvalles – Knox Ave



University Avenue: Front – College



High Volume patches – North of Bloor to York Mills and Finch

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Data Analysis – next steps 

Creation of origin-destination matrix: identify areas that are generating and/or attracting high bicycle volumes - Network Planning



Cycling corridor travel times: identify corridors in which interventions could be implemented to improve cyclist comfort or level of service



Delays at intersections: countermeasures for reducing cyclist wait times (eg. bicycle signals)



Identify streets/routes or corridors with high bicycle presence vs Infrastructure



Filter trips based on the trip purpose, user sociodemographics and cycling preferences

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

App Enhancements 

Heat Map Creation  Add tools for cyclists (high capacity bike parking locations, Staircase channels, tool stations)  Social Media integration?  Option for more alerts, notifications – increase engagement

Goals: 



Cyclists: App becomes an everyday resource to locate cycling infrastructure and cycling services in the city and tool to track personal trip data City: Ongoing source of route choice information to help understand trends and impacts over time and as an aid to network planning

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Next Steps 

Complete analysis of first cut from App data and publish results (2014)



Working on 2015 Data analysis and clean up



Update data in the City’s Open Data web portal



Continue to further identify streets/routes or corridors with high bicycle presence – improvements?



Increase engagement - another contest?

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan

Thank You Questions? Jacquelyn Hayward Gulati [email protected]

More Information: 

www.toronto.ca/cycling - City of Toronto Cycling



www.toronto.ca/cyclingapp - Cycling App



www.toronto.ca/cyclingnetwork - Ten Year Cycling Network Plan



@TO_Cycling

Big Transportation Data for Big Cities – Toronto Cycling App for the Ten Year Cycling Network Plan