by Oliver O’Brien, Urbanist & Researcher, University College- London
Oliver O’Brien is an urbanist and researcher in quantitative geography, at University College London. His specialties include spatial analytics of bikeshare systems across the world, and visualizing socio-economic data such as population demographics.
Many major cities around the world are seeing rapid population growth, resulting in increased strain on existing road and public transportation network infrastructure as the numbers on the daily commute swell. Smart mobility – putting data, information, and options in the hands of the travelling public – has been beneficial to many of these cities, allowing better use of fixed resources and more efficient movement around the urban space. Opening up live and static datasets for public consumption can be inexpensive and straightforward relative to the cost of building new physical infrastructure, particularly where sensor information can be easily accessed through existing control systems and carefully specified new “smart city” infrastructure.
London’s Open Data Portal for Transport
Being based in London, I am fortunate to be a data researcher in a city proactively releasing huge amounts of open data. Transport for London (TfL) is the city administration’s transport manager for many modes (e.g. buses, bikeshare, trams, light rail), operator for some (e.g. metro, cable car, major roads) and regulator for others (e.g. taxis, private hire). It has worked with its divisions and private operators to release large amounts of data, both dynamic (updating live, e.g. metro departure boards and traffic cameras) and static (e.g. infrastructure locations, roadworks and safe taxi lists), as open data, for consumption and augmentation by anyone, including commercial concerns who can create potential business with such data.
The data is available through an Open Data Portal section on its website. Live running and timetable information allows multi-modal journey planners (see below) to be easily created and quickly react to disruption and show alternatives. Visualizations of such data can inform both transport planners and the general public as to the use and operation of transport modes, in the long term and short term respectively. Fine-grained temporal capacity information can be used to encourage changes in travel habits. As both a user of TfL’s transit systems, and a visualizer of its data, I have seen first-hand the benefits of the easy access and utilisation of these datasets. I developed TubeHeartbeat, for example, which uses an open dataset from TfL’s portal, on passenger volumes by quarter-hour, to visualize the short but intense commute periods on London’s “Underground” metro network. I also curated an exhibition “Big Data Here”, which projected live running bus information and traffic camera videos, amongst other hyperlocal open data, onto a screen positioned right by the corresponding bus stop and camera itself.
The Open Data Portal’s datasets also facilitate straightforward and efficient research, for example studying how air pollution impact balances improved health outcomes of a shift to cycling as a commute mode. As both a cyclist and open data researcher, this was a piece of research with both personal and professional interest to me.
Commercial App Eco-Systems
The presence of a reliable, comprehensive, documented, and insightful set of data feeds has resulted in an ecosystem of third-party “apps” being developed for the dominant smartphone platforms (Apple’s iOS and Google’s Android). For example, there are many apps for London’s bikeshare system, most showing a map of availability of bikes or free docking station slots, often augmented with information on nearby cycling infrastructure, a timer/meter indicating the cost of the current hire, etc. Many of the apps adopt a free-at-point-of-use, ad-funded business model, while the more advanced ones cost a small amount to purchase.
A key driver of the success of such eco-systems is the adoption of a stance by the data provider (e.g. the transit authority) that, while there is a value to the data that they release on a free and open basis, the third-party commercial marketplace is better positioned to realize this value, through innovation and different thinking, allowing the transit authority to focus on their core role of running and managing the transport. Additionally, as open data is typically supplied without a service contract, the provision of such data does not place additional support obligations on the operator. A specification on availability of the data, and documentation on its use, is however helpful for rapid commercial adoption. Ultimately, this commercial activity benefits the operator by making mobility smarter, the people using the transit systems, and the economy in general. The best innovations can then typically be incorporated into the operator’s own apps or other information systems. In London’s case, and my anecdotal experience, the app developer community and the final users can be quite vocal where there are data errors or availability outages from the upstream providers.
Development of Standard Open Data Formats and Specifications
There are a number of emerging open data file/syntax formats and data specifications, for smart mobility. In terms of physical file formats, JSON is widely used, with XML also often available. Static open data is often supplied in the CSV or KML file formats, although more sophisticated open data portals combine this with the dynamic data format. For example, TfL’s “Unified” API aims to provide almost all its open data output, dynamic and static, in a consistent JSON form.
GTFS (General Transit Feed Specification) was developed by Google as a standard format for transit timetabling information. Once developed, Google released the specification as an open standard, and encouraged public transport authorities to adopt it, so that their timetable information can be incorporated into the global, multi-modal journey planners, such as Google’s own Google Maps. The designation of the format as open doubtless reassures public authorities concerned with adopting a standard for the benefit of a single corporation. Where large public transport authorities already have established formats, such as the Department for Transport’s TransXChange specification in the U.K., third party companies have worked with either side to provide translation between the formats.
Echoing GTFS, the North American Bike Share Association, a consortium of suppliers and operators of bikeshare systems in the US and Canada, has recently developed and published GBFS (General Bikeshare Feed Specification), which aims to show bike and empty docking space information on a standardized and real-time basis, with a view to its future inclusion in multi-modal journey planners in a similar way to GTFS. GBFS is relatively new but has also started to be adopted and published by some bikeshare operators beyond North America. I am hopeful that this standard will be further adopted in the UK and by other providers, likely to happen quickly should Google Maps or another major journey planner incorporate the US data.
Multi-City Users of Open Data for Smarter Mobility
Adopting the above data standards allows multi-city websites and apps to be created and, just as importantly, maintained, with relative ease, with the upstream data provider being responsible for making changes while maintaining consistency through their API. CityMapper and Google Maps are two large users of open data transit feeds. Both offer multi-modal transport planning across a large number of cities across the world.
It is likely that, as tourists and business visitors explore new cities, they are increasingly using the apps already installed on their smartphones, to move around unfamiliar urban environments in an efficient way. A reduction in out-of-country smartphone network data costs, such as is happening across European Union member states, is likely to drive such habits, and be an additional factor in multi-city apps becoming a key component in enabling smart mobility. Certainly, this has been the case for me, I have typically used both CityMapper and Google Maps for recent work trips and holidays in both America and Europe, rather than needing to fill my smartphone screen with apps designed for just one city.
Integrating Transport with Fare Smartcards
One way to encourage “smart” mobility, where different or multiple modes of transport are used as and when different situations require, as suggested by online journey planners, is the adoption of multi-modal, multi-agency fare smartcards. TfL’s fare smartcard is called Oyster Card, and can be used for almost all forms of public transport in London (including on trains not otherwise controlled by TfL) with some notable exceptions, such as the bikeshare system, which was installed on a drop-in basis rather than being engineered to integrate with Oyster, partly due to cost but also due to the need for a sizeable deposit to be secured when hiring an expensive bikeshare bike. Since the introduction of Oyster Card, TfL has continued to innovate with fare smartcards, now allowing regular “contactless” credit and debit cards to be used. This reduces costs for TfL by passing payment processing administration to a range of card providers that compete which each other and so ensure competitive processing costs that equate to less than the overall cost of administering Oyster Card and its associated provider. In the UK, unfortunately, for commercial and logistical reasons, the emerging smartcard standard outside of London is a different system to Oyster Card. However its take-up has been much slower due to less financial incentives for the user. Ultimately, my prediction is that credit/debit cards will supersede both fare smartcards and season tickets, even for season ticket holders, who would instead receive various types period caps across their long-term use.
In Mexico City, the transport mix is diverse, with private minibuses accounting for a sizeable share of the public transport trips within the city and surrounding area. These services are only lightly regulated and do not offer timetabling, running information or route maps. The city’s transport authority has moved to create a more integrated public transport option by augmenting its very popular but overcrowded metro system with high capacity bus rapid transit (BRT) routes, regular buses, and bikeshare (ECOBICI). The BRT, metro, and city buses use a unified payment system, and are extremely well used, becoming intensely busy during commute periods, both above and below ground. There is obvious scope to increase this as more publicly specified transit becomes available in Mexico City in the future (ITDP, 2017). Mexico has a similar state/city incompatibility with smart cards to the UK that likewise could be mitigated with an adoption of “contactless” credit/debit card use.
Visualizing Taxis and Other Transport in New York City
Private taxi journeys continue to be a key part of a city’s successful smart transit mix, as some journeys require the specific capabilities and locations of private vehicles. The taxi and private hire industry has also innovated, alongside its public transport counterparts. For example, the release of open data for hundreds of millions of cab journeys in New York City has resulted in impressive and dramatic visualizations. Taxi journeys on their own have been mapped, but also combined with public transit vehicle movements across many different modes, to indicate efficiently the spatio-temporal nature of movement around Manhattan and other areas in this large, high-density city.
One of the most successful cab companies, Uber, has been key in the increasing use of maps to help people understand their urban space, by providing a route map of each journey taken by an Uber customer, as part of their resulting receipt. I have personally found that such personalized maps help me understand the geographic layout of places I’ve traveled through on Uber (and other similar cab services), potentially also revealing alternative mobility choices.
Mapping Bike Share
Adoption of standards such as GBFS (above), as well as some of the largest bikeshare operating companies running systems across a large number of cities, has meant that creating a map of live bikeshare information, for hundreds of cities worldwide, has been relatively straightforward. I created the Bike Share Map (O’Brien, 2010) to have an at-a-glance map which immediately tells me what I need to know about the state of my local system. I found it easy to expand the map to additional cities thanks to the adoption of such standards mentioned above. At the time of writing, I have therefore been able to add nearly 200 cities to the map – within each city, a complete map of docking stations, with their empty/full status is shown. Replaying the map for the current and previous day shows flows of bikes, be it the multidirectional flow typically present in cities with a high proportion of tourists, or “tidal” flow seen by commuter-dominated system.
A number of city administrations also release, with the arrangement of their corresponding operators, journey information, indicating typically the start and end docking station name/location and time, for each individual journey. Analysis of such information can reveal additional patterns and characteristics of the usage of bike share, and potentially cycling in general, within each system’s area. For example, my map of best-guess routed flows, based on the release of tens of millions of individual bikeshare journeys on Mexico City’s ECOBICI, reveals the concentration of flows down the city’s major city centre artery, Paseo de la Reforma (O’Brien, 2016), indicating the good use of the dedicated bikeshare lanes that were installed there, certainly born out by personal observation. It also highlights other cycle-heavy areas of the center of the city, suggesting possible areas for potential future cycle infrastructure improvements.
This paper has demonstrated some of the benefits of releasing open data on mobility options in major cities, to better inform users of the mobility services, researchers understanding the city and businesses creating value, insights, and tools from such freely given, valuable datasets.