clm-7223cfb1bd4e495f
quantitative
Between September 2012 and September 2013 a team from the University of Nairobi, MIT and Columbia rode all of Nairobi's 135 matatu routes with GPS-tracking phones
Between 2012 and 2013 a university team rode every one of Nairobi's 135 matatu routes with GPS phones and turned a minibus system with no timetables into a transit feed. The stylised map that followed went viral, the county made it official, and on 26 August 2015 Google Maps began routing passengers by matatu. The team's last feed ran out at the end of 2020; who collects the next one is an open question.
This overview was generated with AI from the public sources listed below and checked against them before publication. It has not had a complete human fact-check — treat it as a starting point and follow the sources.
100 sourced claims on this page, none individually verified yet — turn on “Sources in the text” to see each one.
How this atlas is made · What “last verified” means · What “source-checked” means · The full record behind this page
✎ Reclassified 13 Oct 2026 Status changed: live → completed. · 4 revisions in all What changed →
Part of 1001 Smart Cities — 1,216 documented projects · 🏙️ 4 more in Nairobi · 🚌 Public transport guide · 🚲 Mobility projects
Nairobi moves on matatus, the privately owned minibuses that run its bus routes. The project's peer-reviewed paper put the scale this way: Nairobi's matatu network comprises over 135 routes that, according to the 2009 census, serve a population of well over 3.1 million within the metropolitan area, and it cited the licensing board's count that approximately 9554 matatus and buses serve the Nairobi region. Sarah Williams, who led the project at MIT, told the institute's news office that roughly 3.5 million people depend on matatus every day in Nairobi but have little information on the more than 130 different routes.
What makes the system hard to describe is not its size but its rules. In the paper's words, service does not always have fixed schedules and fares, and drivers often take detours to avoid traffic or police and sometimes take the liberty of improvising stops. Fares follow demand: when it rains in Nairobi, fares can triple. Even the stops are not the city's: the team charted over 3,000 matatu stops, most of which had been established by private matatu drivers, not the local government.
The idea came from a mapping laboratory, not a transport authority. Supported by a grant from The Rockefeller Foundation, the Digital Matatus project was conceived at Columbia University, where Williams worked as director of the Spatial Information Design Laboratory through August 2012, when she moved to MIT. The paper is candid about how new the method was: when our research team started the project in 2012, we did not know of any organizations using mobile devices to generate data on semi-formal bus systems.
The team's own accounts date the fieldwork two ways, and both are true. The paper says data collection occurred from September 2012 to September 2013; however, the process was periodically stopped to test data quality and retool our collection software; the book chapter separates the two phases: in the fall of 2012, we tested various software developed for data collection on cell phones, and formal data collection began in February of 2013 using a modified version of MyTracks App for Android smart phones developed specifically for the project. The riders were a student team from the University of Nairobi led by Dan Orwa and Peter Wagacha.
How many routes they rode depends on which of the team's publications you read. The paper says the team successfully collected data on 135 routes that comprise Nairobi's matatu system; the chapter says we collected 136 routes and 6 rail lines; a Nairobi radio station reported 134 matatu routes and over 3,000 stops in Nairobi County were mapped; Columbia's news office said some 130 matatu routes, covering more than 3,000 stops; and Jacqueline Klopp, one of the project's founders, wrote in 2016 that Cape Town has over 600 routes while Nairobi has 138. The atlas uses the peer-reviewed figure. The spread is not an error so much as a property of the system: routes split, merge and move.
Turning the rides into a transit feed meant bending a standard written for cities with timetables. The General Transit Feed Specification, GTFS, is what Google Maps and most journey planners read, and at the time it barely existed on the continent: a review found that only four of around 766 agencies producing feeds were in Africa. The format, the paper notes, also assumes there are standard schedules and fares, standard vehicle types, scheduled service outages, and that transit agencies are maintaining the data. None of that held. So we had to generate rough estimates for departure and frequency of trips from the main terminus at peak and off-peak periods, on the observation that a matatu generally leaves a stop every two minutes during peak hours. Fares were left out altogether: fare information is optional in the GTFS format; therefore, we decided not to populate this field since it would be difficult to develop it in a standardized way. And because the feed needs an operator, given that the data was developed for the hundreds of "agencies" operating matatus, the research team is listed in that field. The one change the team asked of the standard itself came out of a conference: the group added a "continuous stops" field to the stop times and routes table, for vehicles that stop wherever a passenger waves.
The feed was for machines; the map was for people. In January 2014, Digital Matatus released a stylized transport system map resembling subway maps from developed world capitals like London, New York, and Paris. It was launched with a hackathon: over the weekend of January 25–26, 2014, eight teams of up to four university students each participated in the collaborative programming competition. By the team's account the downloadable maps went viral on social media, and versions were printed in local newspapers.
The number everyone repeats is five thousand. MIT's news office wrote that the map was wildly popular, with over 5,000 downloads registered in the first six months; the project's homepage counts it differently, as more than 5,000 unique IP addresses downloaded the data; the paper says there have been more than 5000 downloads of the data and paper maps. Three descriptions of one figure, all from the team, none of them audited. It is a count of downloads, not of journeys.
Whether the county adopted the map is the first thing the team's own accounts disagree about. The paper states that the Nairobi City County Government has designated the map as the official transit map for the city, and a figure caption says that Nairobi's city government made it the official matatu map when the final edited version was released in January 2014. The chapter says in one place that the governor of Nairobi made it the official Matatu map and, a page later, that the city intended to adopt it as its official map. No county document saying either was found. What is documented is a launch abroad: in October 2015 Governor Evans Kidero was attending the City Lab London 2015 Conference when Digital Matatus was presented as a county system.
The only measurement of what the map did for anyone is a survey the team ran itself. A total of sixty-six surveys were collected after respondents had the opportunity to review and comment on the paper version of the matatu map, from first-year planning students at the Technical University of Kenya. The answers were warm: almost all, 86%, believed the map made it easier to find the way around the city, about 83% reported that with the map they are more likely to take a new matatu route, and more than 80% of respondents reported never having seen a public transit map like the Digital Matatu map for their city. The chapter also says what the sample was not: the survey sample did not reflect the makeup of the city; about two-thirds of respondents were under the age of 22 years of age, and nearly 88% were males.
The step that made the project famous came nineteen months after the map. The paper records that the Google Maps transit feature was launched on August 26, 2015 in Nairobi, and adds the sentence that has been repeated ever since: it is the first semi-formal transit system to be included in the Google Maps transit routing feature. MIT's lab put it the same way, as the first time an informal transit system is navigable through Google Maps.
Google's own announcement, published the same day on its Africa blog, says something narrower. It credits the partnership, working in collaboration with Digital Matatus, and then frames the city as a late arrival: Nairobi now joins the over 2800 towns and cities having this very useful and convenient feature, including Lagos, Johannesburg, Tokyo and Sydney. The word first does not appear. The Nairobi press repeated the team's version, Capital FM calling it a first for Google maps as it will inform the development of Transit maps in similar cities with similar public transport structure like Manila, Dhaka, Mexico City or Dar es Salaam, and the most careful outside phrasing is ITDP's a year earlier: the Nairobi project is one of the first to bring informal transit into GTFS. The atlas attributes the first to the team.
Before Google, the data had already been used. The team assisted Laban Okune, who used the data to improve his award-winning Ma3Route mobile app, and worked with Jeremy Gordon of Flashcast who developed a routing program called Sonar using the data. The chapter adds that the Digital Matatu trip planning app was created for Windows, that another trip planning app called Matatu Map was created, and names a fifth from abroad: more recently a Canadian company used our data to produce Transit App. Planners used it too. ITDP wrote in March 2014 that it was, in a partnership with the Kenya National Highways Authority (KeNHA) and UN-Habitat, building off the dataset to develop plans and recommendations for the upcoming BRT along the Uhuru Highway. The paper's reference list names the document: the Nairobi Ndovu/A104 Bus Service Plan of 2014.
A transit feed is only as good as its last survey, and the team knew it. Klopp put it plainly in 2016: transit systems are dynamic, and it is critical that data is constantly updated. In February 2017 the project announced that an updated version of the Nairobi GTFS Feed is now available on our website, after a second ride through the city in which the next step was to collect updated data for all the 133 matatu routes in Nairobi. MIT's lab page records that in the summer of 2018, the research team performed the third round of data collection, and that the collection software was changed so as to allow a comparison of old routes to new routes and the ability to modify the old routes where needed, which, the lab notes, helped maintain the data in the format needed for Google Maps upload.
The third feed appeared in October 2019. The Digital Matatus team has just released the 2019 transit feed for the city of Nairobi, the announcement said; by then the Digital Matatus team has collected and verified more than 140 routes and 4,000 stops across the city. It called this the third extensive data collection conducted by the team in the past six years and reported that in the past two years, the matatu network has grown from 3,300 km to 3,550 km in length. The project's download page lists the three feeds by date, GTFS Feed (05/30/2019) being the newest, and a 2019 wall map beside them. The October 2019 post is also the newest item on the project's news page.
What the catalogues say about the feed since then is sparse and consistent. The Mobility Database, the successor to the exchange the team once uploaded to, carries two records. The original entry notes that this feed has been replaced with a different producer URL; the replacement is the same producer under a new address, with a service range that reads Start Jan 1, 2019 End Dec 31, 2020, a last download logged as Latest: Fri Jun 07 2024, and a status of inactive. The feed's own calendar, in other words, ran out at the end of 2020, and nothing newer has been published in its place.
Nairobi is not the only city on that new address. The replacing feed is served from the data commons of DigitalTransport4Africa, which describes itself as An Open Resource Center for Mapping Public Transport across Africa and which, in the words of the team's own 2019 paper, was spearheaded by the French Development Agency and Digital Matatus and other mapping collectives and joined by the World Bank. The commons' Nairobi repository describes itself as GTFS, map and papers for the buses of Nairobi (Kenya); by the platform's own metadata it has not changed since April 2022, while the commons itself was still posting in 2026.
The method moved faster than the data. Klopp, writing in 2016, listed the cities that had taken it up: these cities include Kampala, Maputo, Accra, Lusaka, Amman, Cairo, Managua and elsewhere. Kampala was the closest copy: the 2015 paper notes that UN-Habitat and ITDP more recently have tried to replicate this work in Kampala, Uganda, and the 2019 paper that UN-Habitat engaged the Institute of Transportation Development Policy (ITDP) to replicate the mapping process in Kampala. By 2022 the World Resources Institute, which describes the commons as led by WRI and partners and funded by the French Development Agency (AFD), could still write that only 15 African cities have mapped their public transport systems and minibus taxi networks in a standardized and open format.
The city that did what Nairobi did not was Maputo. There, the 2019 paper records, the minibus map was renamed the city's public transport map, a step that indicated the Maputo City Council's willingness to take ownership of the map, a significant milestone in the project. In Nairobi the question of ownership was asked on the day of the Google launch and never answered. At the launch of the Google transit app for Nairobi in 2015, the ICT Director for the National Transport and Safety Authority was asked by journalists who would be responsible for updating and providing this data to the public, the 2019 paper recalls. The team's 2015 paper had already proposed an answer: ideally, this function should eventually be taken over by a government agency, such as Kenya Institute for Public Policy Analysis (KIPPRA) or the National Transport and Safety Authority, with a steady budget allocation for updating the data, adding that KIPPRA has expressed interest in maintaining the data and the methodology. No source read for this page records either agency doing so.
The people who did use the data were, by the founders' own account, the ones least inclined to pay for it. The 2019 paper says that multilateral actors/lenders such as the World Bank, the European Union and the African Development Bank and their consultants, used the data and maps without, however, generally recognising a role in or responsibility for supporting the creation of such data. AFD's own retrospective on the Accra copy, from 2021, reads like a verdict on the whole genre: still poor endorsement by public authorities, and the assumptiun « we have data = we can have relevant planning » is not so obvious, the spelling as in the original.
Nairobi's bus rapid transit system, the plan the data was first borrowed for, is the clearest measure of that gap. ITDP's service plan for Line 2, published in October 2019, begins with passenger demand surveys that help determine how frequently each existing route operates, how many people board and alight at each stop, current commercial speeds, and transfer locations; the document does not mention Digital Matatus or GTFS. In 2026 the corridor the plan was written for is not running. In July the metropolitan transport authority was reported to be seeking funds for a project that was stopped due to lack of funding and planning a scheme that will enable the use of existing public transport routes for 5 corridors; in August the Auditor-General flagged the project as stalled, the contractor having suspended the works from 11 January, 2022. Eleven years after a Nairobi map went into Google Maps, the matatu routes that were mapped are still the routes.
Over 5,000 downloads of the map were registered in its first six months, by the team's own count, and five routing apps were built on the data, among them Ma3Route and Flashcast Sonar in Nairobi and the Canadian Transit App. In the project's own survey of 66 planning students 86% said the map made it easier to find their way around the city and 83% said they were more likely to take a new matatu route. ITDP, with UN-Habitat and the national highways authority, used the data for the 2014 service plan of the Uhuru Highway (A104) bus rapid transit corridor.
A dataset outlives the project that made it only if somebody keeps collecting. The team re-rode the routes in 2017 and 2019 and said the 2019 collection covered more than 140 routes and 4,000 stops; that feed's calendar ran out on 31 December 2020, and the Mobility Database lists it as inactive. No Nairobi agency is recorded in the sources read here as having taken the collection over. What travelled instead was the method - a GTFS loosened for a system where fares vary and routes and stops move with traffic, police checks, commuter demand and the weather.
Two clocks, kept apart: when this entry’s sources were published, and what has happened to its record in the atlas. Documented span: 2012–2020. 18 of 25 sources carry a publication date, from 7 March 2014 to 10 August 2026.
Nearly every statement about the project comes from the project: the peer-reviewed paper and the book chapter were written by the team, the project website is its own, and the two university news offices are partner institutions. The outside voices are Google's announcement, ITDP writing about its own use of the data, two Nairobi news outlets, and a second peer-reviewed paper of 2019 that one of the founders co-wrote. No Nairobi county or national document about the map, the feed or its adoption was found, and no independent study of whether the map changed how anyone travels exists in the sources read here.
Five questions stay open. Whether Google Maps still returns matatu directions for Nairobi in 2026 could not be checked from where this page was written, and no source after 2018 says so. Whether the county formally adopted the map is stated three ways by the team itself. What the figure of 5,000 downloads counted differs between the team's accounts. Which vintage of the feed sits in the DigitalTransport4Africa repository is not readable without downloading it. And of the five routing apps named in 2015, two could be checked: on 11 October 2026 the web domains of Ma3Route and FlashCast did not resolve, which is a sign and not a finding.
Evidence file · public beta 107 addressable claims, 25 source dossiers, one permanent baseline. 100 of them are marked in the text above; the other 7 belong to a passage this page does not print, so they can be read here but not clicked in the article. The file is unreviewed — an AI-assisted structural migration, not a human or expert endorsement. Every revision keeps a permanent copy under revision history.
The text below is copied verbatim from the published record. Each claim names the source dossiers that carry it; caveats stay attached to the claim they qualify.
clm-7223cfb1bd4e495f
quantitative
Between September 2012 and September 2013 a team from the University of Nairobi, MIT and Columbia rode all of Nairobi's 135 matatu routes with GPS-tracking phones
clm-ee57983e7aec5b68
descriptive
converted the data into a modified GTFS loose enough for a system with no calendars, no service frequencies and no operating schedules
clm-2e4fc5252736b2be
quantitative
A stylised paper map followed in January 2014
clm-04f163c7e4a6c34b
descriptive
which the county government made the official matatu map
clm-67e4f3d5a9b323c6
quantitative
on 26 August 2015 Google switched on matatu directions in Google Maps
clm-5df36a8fc61895b3
descriptive
The project team calls it the first semi-formal transit system in Google's routing
clm-c1a28ceabc10766c
quantitative
Google's own announcement said Nairobi joined "the over 2800 towns and cities" with the feature
clm-0549c9974902dc30
quantitative
Over 5,000 downloads of the map were registered in its first six months
clm-dae62774069ee437
quantitative
five routing apps were built on the data, among them Ma3Route and Flashcast Sonar in Nairobi and the Canadian Transit App
clm-895336bd64046262
quantitative
In the project's own survey of 66 planning students 86% said the map made it easier to find their way around the city and 83% said they were more likely to take a new matatu route
clm-6d309dde820a7294
quantitative
ITDP, with UN-Habitat and the national highways authority, used the data for the 2014 service plan of the Uhuru Highway (A104) bus rapid transit corridor
clm-f78c10d3ca86e122
quantitative
The team re-rode the routes in 2017 and 2019
clm-8dcac6f387fc26ba
quantitative
said the 2019 collection covered more than 140 routes and 4,000 stops
clm-d1af6aae39e68aaa
quantitative
that feed's calendar ran out on 31 December 2020, and the Mobility Database lists it as inactive
clm-ce70a1bc7c202ef4
descriptive
No Nairobi agency is recorded in the sources read here as having taken the collection over
clm-a489415038b93f84
descriptive
a GTFS loosened for a system where fares vary and routes and stops move with traffic, police checks, commuter demand and the weather
clm-debd42bdc8a8f1a7
quantitative
Nairobi's matatu network comprises over 135 routes that, according to the 2009 census, serve a population of well over 3.1 million within the metropolitan area
clm-ce53e3f43b1758b8
quantitative
approximately 9554 matatus and buses serve the Nairobi region
clm-a85f9b15d93c873c
quantitative
roughly 3.5 million people depend on matatus every day in Nairobi but have little information on the more than 130 different routes
clm-4d57564c0c47ad6d
descriptive
service does not always have fixed schedules and fares, and drivers often take detours to avoid traffic or police and sometimes take the liberty of improvising stops
clm-012a41cc14b03646
descriptive
when it rains in Nairobi, fares can triple
clm-afc1d787c5293306
quantitative
charted over 3,000 matatu stops, most of which had been established by private matatu drivers, not the local government
clm-07407a36835123f4
descriptive
Supported by a grant from The Rockefeller Foundation, the Digital Matatus project was conceived at Columbia University
clm-9e7102c127c2072d
quantitative
where Williams worked as director of the Spatial Information Design Laboratory through August 2012, when she moved to MIT
clm-2cc3d39c4398194c
quantitative
when our research team started the project in 2012, we did not know of any organizations using mobile devices to generate data on semi-formal bus systems
clm-b25b4eb5ed6b3085
quantitative
data collection occurred from September 2012 to September 2013; however, the process was periodically stopped to test data quality and retool our collection software
clm-b39f8275e0245b1a
quantitative
in the fall of 2012, we tested various software developed for data collection on cell phones
clm-792197e0b6a37d0e
quantitative
formal data collection began in February of 2013 using a modified version of MyTracks App for Android smart phones developed specifically for the project
clm-c06caef9e0c29077
descriptive
a student team from the University of Nairobi led by Dan Orwa and Peter Wagacha
clm-59e23c930f6d48d8
quantitative
successfully collected data on 135 routes that comprise Nairobi's matatu system
clm-045cc0372728dea4
quantitative
we collected 136 routes and 6 rail lines
clm-bc776b1ea9daf6cf
quantitative
134 matatu routes and over 3,000 stops in Nairobi County were mapped
clm-834a8fdeea8dfb36
quantitative
some 130 matatu routes, covering more than 3,000 stops
clm-8d9df7d2e07eb27f
quantitative
Cape Town has over 600 routes while Nairobi has 138
clm-a07983881bc353f2
quantitative
only four of around 766 agencies producing feeds were in Africa
clm-cff82a36b12ab6b7
descriptive
also assumes there are standard schedules and fares, standard vehicle types, scheduled service outages, and that transit agencies are maintaining the data
clm-ce3974236f021759
descriptive
we had to generate rough estimates for departure and frequency of trips from the main terminus at peak and off-peak periods
clm-b1104238b9d4bc14
descriptive
a matatu generally leaves a stop every two minutes during peak hours
clm-ac8063b525387ec9
descriptive
fare information is optional in the GTFS format; therefore, we decided not to populate this field since it would be difficult to develop it in a standardized way
clm-c60ff6ab9fd08235
descriptive
given that the data was developed for the hundreds of "agencies" operating matatus, the research team is listed in that field
clm-0c0594df998758a2
descriptive
the group added a "continuous stops" field to the stop times and routes table
clm-f983715c8131048f
quantitative
In January 2014, Digital Matatus released a stylized transport system map resembling subway maps from developed world capitals like London, New York, and Paris
clm-8e8cde945e44d13c
quantitative
over the weekend of January 25–26, 2014, eight teams of up to four university students each participated in the collaborative programming competition
clm-a17da2fd776f7503
descriptive
the downloadable maps went viral on social media, and versions were printed in local newspapers
clm-a31ecea7558c05d9
quantitative
wildly popular, with over 5,000 downloads registered in the first six months
clm-d48388159065f290
quantitative
more than 5,000 unique IP addresses downloaded the data
clm-6d999f4652caac1e
quantitative
there have been more than 5000 downloads of the data and paper maps
clm-2e43097fda1312f8
descriptive
the Nairobi City County Government has designated the map as the official transit map for the city
clm-e057f99d5b1c020b
quantitative
Nairobi's city government made it the official matatu map when the final edited version was released in January 2014
clm-b29ff2f376e81029
descriptive
the governor of Nairobi made it the official Matatu map
clm-fc32f91d675e8e3c
descriptive
the city intended to adopt it as its official map
clm-be0b7ab46a266d74
quantitative
Governor Evans Kidero was attending the City Lab London 2015 Conference
clm-3b40aff58fdb1d71
descriptive
A total of sixty-six surveys were collected after respondents had the opportunity to review and comment on the paper version of the matatu map
clm-833160659495a584
quantitative
almost all, 86%, believed the map made it easier to find the way around the city
clm-977bd4efd8048f43
quantitative
about 83% reported that with the map they are more likely to take a new matatu route
clm-b07ec3d4b1e0a057
quantitative
more than 80% of respondents reported never having seen a public transit map like the Digital Matatu map for their city
clm-e2ca5d762305c064
descriptive
the survey sample did not reflect the makeup of the city
clm-49a26b9062886930
quantitative
about two-thirds of respondents were under the age of 22 years of age, and nearly 88% were males
clm-92f8895b6ae214ac
quantitative
was launched on August 26, 2015 in Nairobi
clm-e7c5dfd3387077f1
descriptive
it is the first semi-formal transit system to be included in the Google Maps transit routing feature
clm-124f098cd1f880e1
descriptive
the first time an informal transit system is navigable through Google Maps
clm-f197b11805d3e186
descriptive
working in collaboration with Digital Matatus
clm-5f3b5420299e5669
quantitative
Nairobi now joins the over 2800 towns and cities having this very useful and convenient feature, including Lagos, Johannesburg, Tokyo and Sydney
clm-e03910ee1ff8996e
descriptive
a first for Google maps as it will inform the development of Transit maps in similar cities with similar public transport structure like Manila, Dhaka, Mexico City or Dar es Salaam
clm-8ad8622f394b0a4c
descriptive
the Nairobi project is one of the first to bring informal transit into GTFS
clm-65707d85487b29fb
quantitative
The team assisted Laban Okune, who used the data to improve his award-winning Ma3Route mobile app
clm-a8d620abfd019bca
descriptive
worked with Jeremy Gordon of Flashcast who developed a routing program called Sonar using the data
clm-d6120563aa946f64
descriptive
the Digital Matatu trip planning app was created for Windows
clm-db4bd37db15d8238
descriptive
another trip planning app called Matatu Map was created
clm-709e918e16539ed6
descriptive
more recently a Canadian company used our data to produce Transit App
clm-e856aa8dac4eab71
descriptive
in a partnership with the Kenya National Highways Authority (KeNHA) and UN-Habitat
clm-235ff1300e8fb517
descriptive
building off the dataset to develop plans and recommendations for the upcoming BRT along the Uhuru Highway
clm-52c55ca7eca84479
quantitative
Nairobi Ndovu/A104 Bus Service Plan
clm-bcd9525fcc508f6e
descriptive
transit systems are dynamic, and it is critical that data is constantly updated
clm-89530a63f2d25647
descriptive
an updated version of the Nairobi GTFS Feed is now available on our website
clm-c43daf525026de11
quantitative
the next step was to collect updated data for all the 133 matatu routes in Nairobi
clm-9bcd8e9b11369838
quantitative
in the summer of 2018, the research team performed the third round of data collection
clm-54e3e3c346057f3c
descriptive
to allow a comparison of old routes to new routes and the ability to modify the old routes where needed
clm-cf575c52fc69c94c
descriptive
helped maintain the data in the format needed for Google Maps upload
clm-83c7f5001dd21fe4
quantitative
The Digital Matatus team has just released the 2019 transit feed for the city of Nairobi
clm-3caf3c0c9b6dd329
quantitative
the Digital Matatus team has collected and verified more than 140 routes and 4,000 stops across the city
clm-0a13a209d0376485
descriptive
the third extensive data collection conducted by the team in the past six years
clm-a8aa01e8a9352a19
quantitative
in the past two years, the matatu network has grown from 3,300 km to 3,550 km in length
clm-063e99428ad50850
quantitative
GTFS Feed (05/30/2019)
clm-9af04383be90f913
descriptive
this feed has been replaced with a different producer URL
clm-252877199b6a6a80
quantitative
Start Jan 1, 2019 End Dec 31, 2020
clm-9d6956db47d703c3
quantitative
Latest: Fri Jun 07 2024
clm-4766334d7fe0e696
descriptive
An Open Resource Center for Mapping Public Transport across Africa
clm-89d9138496ab96be
descriptive
spearheaded by the French Development Agency and Digital Matatus and other mapping collectives and joined by the World Bank
clm-4785dc7f6e08185b
descriptive
GTFS, map and papers for the buses of Nairobi (Kenya)
clm-1fb16b362fefdddb
descriptive
these cities include Kampala, Maputo, Accra, Lusaka, Amman, Cairo, Managua and elsewhere
clm-3f7c958f9e2e8fe9
descriptive
UN-Habitat and ITDP more recently have tried to replicate this work in Kampala, Uganda
clm-b96b2a986ba2b6c9
descriptive
engaged the Institute of Transportation Development Policy (ITDP) to replicate the mapping process in Kampala
clm-f4a039007fb2e8a1
descriptive
led by WRI and partners and funded by the French Development Agency (AFD)
clm-e821ffec404ba329
quantitative
only 15 African cities have mapped their public transport systems and minibus taxi networks in a standardized and open format
clm-cd9236a9572a5dd9
descriptive
indicated the Maputo City Council's willingness to take ownership of the map, a significant milestone in the project
clm-90999d4e1ae892f7
quantitative
At the launch of the Google transit app for Nairobi in 2015, the ICT Director for the National Transport and Safety Authority was asked by journalists who would be responsible for updating and providing this data to the public
clm-2cc1fdf6c425cac5
descriptive
ideally, this function should eventually be taken over by a government agency, such as Kenya Institute for Public Policy Analysis (KIPPRA) or the National Transport and Safety Authority, with a steady budget allocation for updating the data
clm-edeea7128a36a879
descriptive
KIPPRA has expressed interest in maintaining the data and the methodology
clm-5a09972ad855a687
descriptive
multilateral actors/lenders such as the World Bank, the European Union and the African Development Bank and their consultants, used the data and maps without, however, generally recognising a role in or responsibility for supporting the creation of such data
clm-c3c41e2a3a843cff
descriptive
still poor endorsement by public authorities
clm-7e3646045ec1675f
descriptive
the assumptiun « we have data = we can have relevant planning » is not so obvious
clm-86a2ed701bae92aa
descriptive
begins with passenger demand surveys that help determine how frequently each existing route operates, how many people board and alight at each stop, current commercial speeds, and transfer locations
clm-be11f866bc9c07a0
descriptive
was stopped due to lack of funding
clm-5f147be7afcb9f95
quantitative
will enable the use of existing public transport routes for 5 corridors
clm-00436ef42c1cd8b1
descriptive
the Auditor-General flagged the project as stalled
clm-c9ade84374665b99
quantitative
suspended the works from 11 January, 2022
This layer records what is known about each source, what is still missing, and the exact snapshot this page is rendering. A missing date, archive or hash stays visible as missing.
Awaiting operator and domain-expert review; publication of this revision implies neither.
sha256:7a21a8f72ca0f1345dedde65f7346954d43fdd9746466d0eca78574f88595dac
src-af044b35a4cfe32fThe project's own account of the method, the modified GTFS, the download count, the five routing apps and the city's recognition of the map. A standing project site with no publication date; its newest news item is from October 2019.
src-e21c8f75040e7daeLists the pocket and wall maps dated 23 January 2014, the GTFS feeds dated 5 August 2015, 30 January 2017 and 30 May 2019, and the 2019 wall map. A standing page with no publication date.
src-4d23db754ea14b82The team's announcement of the 2019 feed: data collected since 2014, the third extensive collection in six years, more than 140 routes and 4,000 stops, and a network grown from 3,300 km to 3,550 km in two years.
src-726311a5be7f4c6fAnnounces the first update of the feed, collected for all 133 matatu routes, with upload to Google Maps to follow.
src-d29dac99200786c2The lead laboratory's project page: the January 2014 release, two updates to the GTFS file after the first release, a third round of data collection in the summer of 2018, and the funders. A standing page with no publication date.
src-6db181e6e4bf0b54The lab's news item dating Google's announcement to 26 August 2015 and describing it as the first time an informal transit system is navigable through Google Maps - the team's own wording.
src-edb76fac1428b626Dates the paper map to January 2014, reports 5,000 downloads in its first six months, data collection from February 2013, more than 3,000 stops, and five apps using the data, and describes the Google Maps integration as a first for a semiformal system.
src-c87c43158dc2808fThe peer-reviewed account: data collection from September 2012 to September 2013 on all 135 routes, the modified GTFS, the county's designation of the map as official, the launch of Google Maps transit on 26 August 2015, and the ITDP/UN-Habitat bus service plan. Publication date from Crossref; the publisher blocks automated reading.
clm-7223cfb1bd4e495fclm-ee57983e7aec5b68clm-04f163c7e4a6c34bclm-67e4f3d5a9b323c6clm-5df36a8fc61895b3clm-dae62774069ee437clm-6d309dde820a7294clm-a489415038b93f84clm-debd42bdc8a8f1a7clm-ce53e3f43b1758b8clm-4d57564c0c47ad6dclm-012a41cc14b03646clm-07407a36835123f4clm-2cc3d39c4398194cclm-b25b4eb5ed6b3085clm-59e23c930f6d48d8clm-a07983881bc353f2clm-cff82a36b12ab6b7clm-ce3974236f021759clm-b1104238b9d4bc14clm-ac8063b525387ec9clm-c60ff6ab9fd08235clm-0c0594df998758a2clm-8e8cde945e44d13cclm-6d999f4652caac1eclm-2e43097fda1312f8clm-e057f99d5b1c020bclm-92f8895b6ae214acclm-e7c5dfd3387077f1clm-65707d85487b29fbclm-a8d620abfd019bcaclm-52c55ca7eca84479clm-3f7c958f9e2e8fe9clm-2cc1fdf6c425cac5clm-edeea7128a36a879src-325a89732ef15bf4The team's chapter in Planning Support Systems and Smart Cities: the MyTracks collection from February 2013, 136 routes and 6 rail lines, the survey of 66 planning students (86% and 83%), and the ITDP bus service plan. Online date from Crossref (DOI 10.1007/978-3-319-18368-8_12); the PDF on the project site.
clm-04f163c7e4a6c34bclm-dae62774069ee437clm-895336bd64046262clm-b39f8275e0245b1aclm-792197e0b6a37d0eclm-045cc0372728dea4clm-a17da2fd776f7503clm-b29ff2f376e81029clm-fc32f91d675e8e3cclm-3b40aff58fdb1d71clm-833160659495a584clm-977bd4efd8048f43clm-b07ec3d4b1e0a057clm-e2ca5d762305c064clm-49a26b9062886930clm-d6120563aa946f64clm-db4bd37db15d8238clm-709e918e16539ed6src-ed8c8fc320efa07cThe catalogue record of the original feed: service range 2 March 2012 to 31 December 2015, last fetched 5 May 2016, and the notice that the feed has been replaced with a different producer URL. A live catalogue page with no publication date.
src-a6c4d7a13679ae42The replacing record: the same producer, DigitalMatatus, now served from the DigitalTransport4Africa repository - 136 bus routes, service range 1 January 2019 to 31 December 2020, last downloaded 7 June 2024, status inactive. A live catalogue page with no publication date.
src-ff08226861c488fbGoogle's own announcement: matatu routes available on maps.google.co.ke in collaboration with Digital Matatus, with Nairobi joining 'the over 2800 towns and cities' that have the feature. Google does not call it a first.
src-4dfb4e26abb551bcITDP's own account of using the dataset, in partnership with the Kenya National Highways Authority and UN-Habitat, for plans and recommendations for the BRT along the Uhuru Highway; calls Nairobi one of the first projects to bring informal transit into GTFS.
src-e3822a160264c8d7Nairobi press report of the launch day: 134 routes and over 3,000 stops mapped, and the integration described as a first for Google Maps.
src-fdb11e71d2ddd81dThe partner university's news office: some 130 routes and more than 3,000 stops, and the map downloaded more than 5,000 times in its first six months.
src-6e6f1383a52d4e22Reports Governor Evans Kidero launching Digital Matatus as a county system at the CityLab London 2015 conference and repeating that the data was used by five apps in Nairobi.
src-93f6f79d39e2115fKlopp and Cavoli, Transport Reviews 39(5): the spread of the method to other African cities, the Maputo council's ownership of its map, the unanswered question of who updates Nairobi's data, and the lenders who used it. Online date from Crossref; read from the UCL open-access copy.
src-183bd1e7793f2609Jacqueline Klopp's own account a year after the Google launch: the cities taking up the method, Nairobi's 138 routes against Cape Town's 600, and the need for constant updating.
src-0432d09f4d0121faWRI's release: DigitalTransport4Africa is led by WRI and partners and funded by AFD, and only 15 African cities had mapped their public transport in a standardised open format by mid-2022.
src-9e0e6444d5dc011aThe resource centre's homepage: its self-description, management by WRI and AFD, and community posts dated into 2026. A standing page with no publication date.
src-921c4a123b371d40The commons' Nairobi repository: GTFS, maps and papers; creation and last activity both dated 25 April 2022 in the platform's metadata. The producer URL of the Mobility Database's replacing feed points here. A live repository with no publication date.
src-52622766854b125fAFD's slide deck on paratransit mapping and digital commons: the Accra copy's output and the agency's own lessons, including poor endorsement by public authorities. Dated on the slides.
src-d24a04650c3af437ITDP Africa's service plan for Nairobi's BRT Line 2: the method rests on the plan's own passenger surveys; the text does not mention Digital Matatus or GTFS. Dated October 2019 in the document.
src-81e9232e2e772169Capital FM on the Auditor-General's report for the year to 30 June 2025: the Thika Road BRT works suspended from 11 January 2022 and the project flagged as stalled.
src-74de368b58d8937dThe Star on the metropolitan transport authority's plans in July 2026: the BRT project stopped for lack of funding and a scheme to use existing public transport routes on five corridors.
1001 Smart Cities (2026). “Digital Matatus: Nairobi's Matatu Routes Mapped into GTFS” — Nairobi, Kenya. The Smart City Atlas. https://1001smartcities.org/projects/nairobi-digital-matatus/ (last verified 2026-10-11). Data: CC BY 4.0.
The underlying data is free to reuse with attribution — see the open data page.