OpenStreetMap is an international community of citizens who are experts in open geographical data, from different perspectives and specialties, and in methods of collaborative construction of information. OpenStreetMap is part of the trend of Voluntary Geographic Information.
The Mexican community of OpenStreetMap is mainly involved in 3 lines of action:
- Strengthening the community (training and methodological support in projects, integration with academia and activism)
- Procedures with government institutions for the importation of large public databases
- Management with institutions to promote open and collaborative practices in its information management
As a community and ecosystem of technologies, OpenStreetMap has participated last years in several initiatives of participatory mapping of informal transport. This lack of good transport systems and lack of data is crucial in developing cities.
Objectives of transport mapping initiatives
These operations usually pretend to:
- Create a complete transport system map to allow citizens to orient themselves in their city and make their mobility more efficient. Having this map allows more people to use collective transport instead of individual transport, with important implications on the mobility system, economy, environmental conditions and the quality of life of people
- Provide local governments with data to regulate their transport, when they do not exist in useful formats
- Provide city and transportation planners with the necessary information to generate diagnostics, analysis, and design
- Allow research and innovation on the subject of transport and mobility with georeferenced open data and with quality standards
In order to make it possible for organizations, communities and governments to systematize the different experiences already developed in the world, the OpenStreetMap Mexico group undertook a medium-term exercise in collaboration with WRI and MIT to document all cases, evaluating, and highlighting successful practices in terms of organization, management, participatory method, technology, data produced, socialization and outputs given to the data, as well as the methods used to be able to update them. The result may be in the future a platform that catalyzes and shares knowledge: where in addition to providing the open technologies and data generated in different cities, a methodological guide is shared with the methods that have led to the best results.
The experiences developed in the world - Mapanica in Managua, Matatus in Nairobi for the best known, but also in Egypt, India, Bangladesh, Colombia, Mexico, etc. Have been led by different organizations, usually in an horizontal collaboration: associations, NGOs, communities such as OpenStreetMap, groups of civic technologies, universities, small businesses or startups, local governments.
This text is the first step of an evaluation work line of participatory transport mapping operations. This first evaluation is about the final data of MapatonCDMX developed in Mexico City.
Guidelines on the work process of MapatónCDMX
The results presented are from the MapatonCDMX organized in 2015-2016 by the “Laboratorio para la Ciudad”, the area of innovation of the Government of Mexico City. The Laboratory proposed to develop a methodology and participatory dynamics to carry out mapping on a voluntary basis.
The MapatonCDMX was initially developed based on a working method applied before in Mexico City and other countries (Bangladesh among others) by the organization Urban Launchpad with the app Flock Tracker. The strategy changed in the course to incline towards the development of an application proprietary for the track of the transport routes, with essentially the same methodology.
In parallel, a collaborative table was formed with participants from various profiles and institutions. The OpenStreetMap group had the opportunity to participate in some working meetings where we suggested the use of existing open technologies instead of developing an app from zero, and to provide in the information structure the necessary attributes to the later generation of GTFS **.
The participatory dynamics was directed towards methods of gamification, based on an open call and the formation of teams put in competition for the obtaining of sponsors’ prices. To improve participation, other groups of young people already enrolled in a voluntary process (INJUVE youth) were added to the dynamic. The organizations represented at the collaborative table were also asked to contribute with the provision of volunteers.
The data collection season resulted in a data set that was subsequently cleaned with the contribution of several organisms, in 2 stages, with a relative result.
MapatonCDMX data evaluation
The data of the MapatonCDMX were published in different formats, the files of the GTFS specification can be downloaded here, the complete database can be consulted from its dashboard and API for developers.
The first public delivery of data via the API, yielded a route system with topology deficiencies and an erroneous validation process.
Photo 1. Routes contained in the API of the MapatonCDMX in the first delivery
In the Mapaton API, which was the method for public consultation and download of routes in GTFS format in the first delivery, 2893 routes were available, of which only 501 were valid for GTFS format, representing only 20% usable in navigation and network analysis technologies.
Photo 2. Routes contained in the GTFS in the first delivery
The routes contained in the GTFS, despite having a good topology, did not manage to be representative of the public transport offer in the CDMX. The state of the information requires a process of validation and correction of the routes contained in the API that could not be integrated to the GTFS because of the topological errors.
The data were submitted to a second cleaning session with the support of the Mario Molina Center in October and November 2016. These data were made available in December. In this second stage, the API presents 4019 routes, of which 29 do not contain geographical information, the 3990 remaining routes are divided into categories:
- Valid: 2162
- Invalid: 1122
- Invalid, out of time travel: 529
- Invalid, data with error: 177
Of these routes, 501 were finally added in the GTFS format, which represents 23.17% of the total of API’s valid routes. For these 501 routes, georeferencing is good and is attached to the road network, but some of the attribute information is missing and does not allow to validate the format and give it a possible use: service schedule, stops, time of arrival at stops. When this information can not be reliably robust (such as in the case of informal or unstructured transport systems mapping), the construction of a GTFS format relies on estimations, but information can not be dispensed with.
A visual summary of the information contained in the GTFS is presented:
Photo 3. Routes total Photo 4. Valid routes Photo 5. Invalid routes Photo 6. Invalid routes, short travels Photo 7. Invalid routes, travels out of time
As a synthesis
The current MapatonCDMX data has a very low part of valid strokes and the absence of the complementary attributes doesn’t allow its use as a GTFS. For this purpose, this data may still go through an in-depth process of validation and correction, which could be done with some technological tool or with a crowdsourcing methodology, which should be submitted to a cost-benefit evaluation. Another strategy could be to re-raise the data or establish a permanent mechanism for contributing to the create new data over a long period, to progressively replace non-functional information.
Because of the scope it has had, the MapatonCDMX can be seen as a pilot test of cooperation and methods of citizen participation. We invite you to read soon about the methods and results of other non-structured transport mapping operations carried out in other cities around the world. We will inform about the upcoming evaluations in this blog and networks of OpenStreetMap México.
** The General Transit Feed Specification (GTFS) was created in 2005 by Trimet and Google so that transit agencies could share open information in a simple way, boost the technology development ecosystem and facilitate the development of applications that use this information.
Additional resources for data visualization:
By @imleodc & @mapeadora