Mapping the geographies of travel using volunteered app data
Emerging data collection methodologies have the potential to enrich knowledge around how people traverse city-regions and their subjective experiences thereof. The July 2021 Map of the Month uses information gathered from volunteers to gain insights into their mobility patterns across Gauteng as well as their associated experiences.
In 2016, GCRO partnered with Lindsay Blair Howe, a research associate at ETH Zürich, on the ‘myJoziMoves’ project, which tracked participants’ movements through Gauteng, using innovative mobile phone technology. The myJoziMoves application collected detailed trip data for individual users' daily commuting across Gauteng (Howe, 2021).
The July 2021 Map of the Month draws on the myJoziMoves dataset, which includes trips from 368 users, captured between October 2016 and January 2017. Each user completed a short set of demographic questions and a number of questions related to their personal experiences of traveling within Gauteng. Each user gave explicit permission for their movements to be tracked by the application, which collected travel data that is geographically precise and accurate with regards to the mode of transport.
Map 1 presents the complete dataset, including over 600 000 data points. Although this was not a representative sample of Gauteng residents, the light grey ‘trails’ demonstrate the extent of day-to-day travel that takes place across Gauteng. The trips of two participants have been highlighted within the map - one a participant with low monthly household income and one with high monthly household income (See Table 1 for demographic details for each). The data from these two participants shows that they have a large geographic footprint as both participants originate from the periphery and travel into the center of the city-region. These participants commute from their homes in Atteridgeville and Morehill (Benoni), respectively. The data also shows that the two participants cross paths somewhere along their travels, despite using different modes of transportation (as can be seen in Table 1).
Map 1 - Volunteer generated movement data from the myJoziMoves application, featuring the trips of two participants
Table 1 - Featured respondents
Map 2 highlights favourite and avoided places as reported by users. Where detail regarding the location or place was limited, mentions were collapsed into subplaces, while general responses were excluded during data cleaning, to ensure standardisation. The map shows avoided places in red and favourite places in green. Of the subset of usable data points, participants identified 43 places across the province as favourites, while 27 were identified as avoided places. There was a high density of favourite and avoided places in Johannesburg, including its northern suburbs and surroundings, followed by the Pretoria CBD and its immediate surroundings. Johannesburg CBD was mentioned 26 times as a favourite place, followed by Rosebank (13 mentions), Sandton and Braamfontein (each with 7 mentions) and Melville, Pretoria, Johannesburg South and Soweto (each with 5 mentions). Interestingly, some places across Gauteng were mentioned as both a favourite place and an avoided place. Johannesburg was mentioned 43 times as an avoided place, followed by Sandton (15 mentions) and Hillbrow (14 mentions).
Closer observation of the data shows that favourite places were chosen because of associations with enjoyable social activities, and opportunities to spend time with friends and family. In addition, places where participants could shop, places of employment as well as places where a range of activities are concentrated within an area were identified by participants as favourites. People tend to avoid places because of perceived crime and safety concerns, and travel-related barriers such as traffic, distance and cost. These reflections highlight the mixed attitudes and experiences of Johannesburg. On the one hand it is a city synonymous with employment opportunities as well as allowing diversity and interaction with people from different socioeconomic backgrounds. On the other hand, it is a place some would rather avoid.
Map 2 - Favourite and avoided places in Gauteng
Map 3 draws straight lines between participants' homes and their favourite places. The map reveals that participants’ favourite places are not necessarily close to where they live, and in some instances people travel significant distances to visit their favourite place. Distances between home and favourite place ranged from 1 km to 64 km, with an average of 18 km. For example, there is a participant who lives in Leacheville (southeast of Benoni) and has listed Bronkhorstspruit as their favourite place, which is some 63 km away from their place of residence. However, in reality, the travel distances along roads or transport routes would be much longer than these straight line distances.
Map 3 - Distance between favourite place and home
These maps show the potential of volunteer generated data for getting a clearer understanding of travel patterns and choices across the province. This type of data provides detailed insight into mobility patterns and could be valuable in informing transport and settlement planning.
While this Map of the Month only shows some of the initial work that has been done using the MyjoziMoves data, the data has also been used in a study which looked at how movement is used as a strategy and resource by those living in poverty to navigate spatial inequality (Howe 2021). The level of detail available in the dataset also enables the analysis of detailed travel patterns by gender, population group and income. It also provides an opportunity to analyse the data spatially and non-spatially based on the reasons for why people avoid places or choose them as their favourites.
Howe, L. B. (2021). Thinking through people: The potential of volunteered geographic information for mobility and urban studies. Urban Studies (online first).