VeriXiv VeriXiv 3029-0988 F1000 Research Limited London, UK 10.12688/verixiv.989.1 Software Tool Article Articles traveltime: an R package to calculate travel time across a landscape from user-specified locations

[version 1; peer review: 1 approved, 2 approved with reservations]

 Ryan Gerard E. Conceptualization Formal Analysis Investigation Methodology Software Visualization Writing – Original Draft Preparation Writing – Review & Editing https://orcid.org/0000-0003-0183-7630 a 1 2 Tierney Nicholas Methodology Software Validation Writing – Review & Editing 1 3 Golding Nick Conceptualization Funding Acquisition Supervision Writing – Review & Editing 1 4 Weiss Daniel J. Conceptualization Methodology Software Writing – Review & Editing 1 3 The Kids Research Institute Australia, Nedlands, WA, 6009, Australia The University of Melbourne School of Population and Global Health, Melbourne, Victoria, 3010, Australia Curtin University, Bentley, WA, Australia The University of Western Australia Department of Mathematics and Statistics, Perth, Western Australia, Australia gerry.ryan@thekids.org.au

No competing interests were disclosed.

 21 5 2025 2025 2 99 15 5 2025 Copyright: © 2025 Ryan GE et al. 2025 This is an open access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Understanding and mapping the time to travel among locations is useful for many activities from urban planning to public health and myriad others. Here we present a software package — traveltime — written in and for the language R. traveltime enables a user to create a raster map of the travel time over an area of interest from a user-specified set of locations defined by geographic coordinates. The result is a raster of the area of interest where the value in each cell is the lowest travel time in minutes to the nearest of the supplied locations. We envisage this software having diverse applications including: estimating sampling bias, allocating defibrillators, setting health districts, or mapping access to vehicle chargers and agricultural facilities. The work-flow requires two key steps: preparing a friction surface for the area of interest, and then calculating travel time over that surface for the points of interest. traveltime is available from R-Universe and GitHub, and documented at https://idem-lab.github.io/traveltime/.

 R geographic information systems spatial analysis Gates Foundation INV-021972 This work was supported, in whole or in part, by the Gates Foundation [INV-021972]. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Introduction

Understanding and mapping the time to travel among locations is useful for many activities from urban planning ( Zahavi 1974) to public health ( Hulland et al. 2019; Weiss et al. 2020) and myriad others ( Nelson et al. 2019). Global maps of travel time to cities ( Weiss et al. 2018; Nelson et al. 2019) and health care ( Hulland et al. 2019; Weiss et al. 2020) have generated much interest and use 1 , and the city data set of Nelson et al. (2019) is available to R users through the widely-used geodata package ( Hijmans et al. 2024). Here we extend that work to enable travel time calculations from any arbitrary set of locations and friction surface.

We present a software package — traveltime — written in and for the language R ( R Core Team 2024). traveltime enables a user to create a raster map of the travel time over an area of interest from a user-specified set of locations defined by geographic coordinates. The result is a raster of the area of interest where the value in each cell is the lowest travel time in minutes to the nearest of the supplied locations.

A gaggle of R packages provide superficially similar though fundamentally different functionality via the TravelTime.com API ( TravelTime 2024a, 2024b; von Bergmann 2024; Lo Russo 2024). Their ‘Isochron’ polygons — areas reachable within a given time from a given location — are most comparable to what traveltime::calculate_travel_time() calculates. However, each isochron is a single polygon calculated is for a single point location and specified maximum travel time, rather than a raster of temporal gradation across a landscape, jointly for an arbitrary number of points, as in traveltime. TravelTime.com cannot provide a single result surface for time to the nearest of a group of points, and continuous time scale without extensive repeated iteration for all combinations of time and points, plus additional calculation of the minimum value for each cell from all points. Furthermore, TravelTime.com requires access keys, a paid subscription beyond a short free period, and caps queries, which add considerable friction to the use of this resource.

With traveltime, we provide free and open source software to estimate travel time from any number of user-supplied locations, across a complete area of interest, and with convenient access to motorised transport or walking friction surfaces with global coverage.

 Methods Implementation

traveltime is an R ( R Core Team 2024) package and requires installation of R version 4.1 or a more recent version traveltime provides a spatial interface using object classes from the terra package ( Hijmans 2024). Travel time is calculated as movement over a resistance ‘friction surface’ ( van Etten 2017). To provide easy access to the existing friction surfaces generated by Weiss et al. (2020), traveltime internally uses the R package malariaAtlas ( Pfeffer et al. 2018) to download surfaces for the area of interest; though users can also supply any other friction surface raster.

 Operation

The work-flow requires two key steps:

preparing a friction surface for the area of interest, and then

calculating travel time over that surface for the points of interest.

Installation

traveltime is available from R-Universe and GitHub, and documented at https://idem-lab.github.io/traveltime/. It can be installed in R as follows: install.packages ( "traveltime" , repos = c ( "https://idem-lab.r-universe.dev" ))

Example: walking from public transport in Singapore

Here we provide an example to calculate and map the walking travel time from rail transport stations across Singapore. Complete code for this example is available as a vignettte in package documentation.

Prepare data and friction surface

We need two items of data:

our area of interest — Singapore, and

our points to calculate travel time from — Singapore’s Mass Rapid Transit (MRT) and Light Rail Transit (LRT) stations.

We download singapore_shapefile, a polygon boundary of Singapore from the GADM ( GADM 2022) database using the geodata package ( Hijmans et al. 2024) as our area of interest: library (terra) library (geodata) singapore_shapefile <- gadm ( country = "Singapore" , level = 0 , path = tempdir (), resolution = 2 ) singapore_shapefile class : SpatVector geometry : polygons dimensions : 1, 2 (geometries, attributes) extent : 103.6091, 104.0858, 1.1664, 1.4714 (xmin, xmax, ymin, ymax) coord. ref. : lon/lat WGS 84 (EPSG:4326) names : GID_0 COUNTRY type : <chr> <chr> values : SGP Singapore

Next we use the function traveltime::get_friction_surface to retrieve a walking friction surface for our area of interest. Alternatively, we could use surface = "motor2020" for motorised travel. We’re also only interested in walking on land so we mask out areas outside of the land boundary in singapore_shapefile. Users supplying their own friction surfaces do not need to download one in this fashion, only ensure that it is in SpatRaster format. friction_singapore <- traveltime :: get_friction_surface ( surface = "walk2020" , extent = singapore_shapefile ) |> terra :: mask (singapore_shapefile) <GMLEnvelope> |-- lowerCorner: 1.1664 103.6091 |-- upperCorner: 1.4714 104.0858Start tag expected, '<' not found friction_singapore class : SpatRaster dimensions : 37, 57, 1 (nrow, ncol, nlyr) resolution : 0.008333333, 0.008333333 (x, y) extent : 103.6083, 104.0833, 1.166667, 1.475 (xmin, xmax, ymin, ymax) coord. ref. : lon/lat WGS 84 (EPSG:4326) source(s) : memory varname : Accessibility__202001_Global_Walking_Only_Friction_Surface_1.1664,103.6091,1.4714,104.0858 name : friction_surface min value : 0.01200000 max value : 0.06192715

Our points are the traveltime::stations data, containing coordinates of all LRT and MRT station exits in Singapore ( Land Transport Authority 2019): library (traveltime) head (stations) x y [1,] 103.9091 1.334922 [2,] 103.9335 1.336555 [3,] 103.8493 1.297699 [4,] 103.8508 1.299195 [5,] 103.9094 1.335311 [6,] 103.9389 1.344999

We plot these data below. traveltime takes resistance values of friction ( van Etten 2017), so higher values of friction indicate more time travelling across a given cell.

Calculate and plot the travel time

With all the data collected, the function calculate_travel_time() takes the friction surface friction_singapore and the points of interest in stations, and returns a SpatRaster of walking time in minutes to each cell from the nearest station: travel_time_singapore <- calculate_travel_time ( friction_surface = friction_singapore, points = stations ) travel_time_singapore class : SpatRaster dimensions : 37, 57, 1 (nrow, ncol, nlyr) resolution : 0.008333333, 0.008333333 (x, y) extent : 103.6083, 104.0833, 1.166667, 1.475 (xmin, xmax, ymin, ymax) coord. ref. : source(s) : memory name : travel_time min value : 0 max value : Inf

We present the resulting calculated travel times in Figure 2 where, as expected, the travel times are lowest near station exits (per Figure 1) and progressively higher further away. Note that the results in travel_time_singapore include infinite values ( Inf above). In Figure 1, the islands to the south and north-east are shown as filled cells, but unconnected with the mainland. The raster cells for these islands appear absent in Figure 2. Because they are not connected to any cells with a station, the calculated travel time is infinite, and so these cells do not appear in Figure 2.

Friction surface raster of Singapore, showing Singapore boundary in grey, and station locations as grey points. Map of walking travel time in Singapore, in minutes from nearest MRT or LRT station. Discussion

The traveltime package is immediately suitable to be used ‘out-of-the-box’ with any set of coordinates, in any part of the globe. We envisage this software having diverse applications including: estimating sampling bias ( Dennis and Thomas 2000), allocating defibrillators ( Tierney et al. 2018), setting health districts ( Padgham et al. 2019), or mapping access to vehicle chargers ( Falchetta and Noussan 2021) and agricultural facilities ( Zhao et al. 2023). Nonetheless, we see opportunities to build the package utility through:

capability to better distribute a wider range friction surfaces, and

additional methods to efficiently compute results over large spatial extents.

Firstly, traveltime currently facilitates access to walking and motorised friction surfaces for 2020, both at 30 arc-second resolution (approximately 0.008333 decimal degrees, or just below 1 km 2 at the equator). Although the user can presently supply their own friction surface, we expect most applications will use these existing surfaces given the extensive work needed in creating new ones ( Weiss et al. 2018, 2020). As landscapes are dynamic, it may be useful to incorporate updated versions of these friction surfaces if and when they are available. Furthermore, although the resolution of these data is likely to be suitable for larger landscape foci, higher resolution data may be helpful for more locally focussed analyses. For instance, although the example here was chosen for it’s simplicity and low computational demands, a ~1 km 2 cell size is a relatively large area to walk across, and thus actual waking times may vary significantly within each cell. We underline however that a user can provide their own higher resolution friction surface at present if desired.

At the other end of the scale, as the area of interest increases, the size of the matrix of calculations necessary increases exponentially, making significant memory demands for analyses over large landscapes (e.g. analyses over Africa required ~ 72 GB RAM to run successfully). Developing methods to handle large landscapes either with less memory or via cloud resources would be helpful to make such analyses accessible to those without access to larger computing resources.

Lastly, although this article is intended to be the key reference for the traveltime package, we suggest citations of the package are accompanied by citing the underlying methodological work as well ( Weiss et al. 2018, 2020).

 Figure permissions

The authors confirm ownership of the figures used in this manuscript.

 Data and software availability

All software described here is available from R-Universe https://idem-lab.r-universe.dev/traveltime and GitHub https://github.com/idem-lab/traveltime, and documented at https://idem-lab.github.io/traveltime/. Code used in this paper is released via GitHub as v4, and archived on Zenodo under DOI 10.5281/zenodo.15347016.

The package associated with this paper contains information from the dataset “LTA MRT Station Exit (GEOJSON)” accessed on the 10th of December 2024 from data.gov.sg, which is made available under the terms of the Singapore Open Data Licence version 1.0 https://data.gov.sg/open-data-licence.

 Acknowledgements

An earlier version of this article has been archived on EarthArXiv https://doi.org/10.31223/X56M74.

 References Dennis RLH Thomas CD : Bias in Butterfly Distribution Maps: The Influence of Hot Spots and Recorder’s Home Range. J. Insect Conserv. 2000;4:73–77. 10.1023/A:1009690919835 Falchetta G Noussan M : Electric Vehicle Charging Network in Europe: An Accessibility and Deployment Trends Analysis. Transp. Res. Part D: Transp. Environ. 2021;94:102813. 10.1016/j.trd.2021.102813 GADM: GADM Maps and Data. 2022. Reference Source Hijmans RJ : terra: Spatial Data Analysis. 2024. Reference Source Hijmans RJ Barbosa M Ghosh A : geodata: Download Geographic Data. 2024. Reference Source Hulland EN Wiens KE Shirude S : Travel Time to Health Facilities in Areas of Outbreak Potential: Maps for Guiding Local Preparedness and Response. BMC Med. 2019;17:1–16. 10.1186/s12916-019-1459-6 Land Transport Authority: LTA MRT Station Exit (GEOJSON) Dataset. 2019. Reference Source Reference Source Lo Russo T : traveltime - a Traveltime API Wrapper for R. 2024. Reference Source Nelson A Weiss DJ Etten J van : A Suite of Global Accessibility Indicators. Scientific Data. 2019;6(1):266. 31700070 10.1038/s41597-019-0265-5 PMC6838165 Padgham M Boeing G Cooley D : An Introduction to Software Tools, Data, and Services for Geospatial Analysis of Stroke Services. Front. Neurol. 2019;10:743. 31440197 10.3389/fneur.2019.00743 PMC6693386 Pfeffer DA Lucas TCD May D : malariaAtlas: An R Interface to Global Malariometric Data Hosted by the Malaria Atlas Project. Malar. J. 2018;17:1–10. R Core Team: R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing;2024. Reference Source Tierney NJ Jost Reinhold H Mira A : Novel Relocation Methods for Automatic External Defibrillator Improve Out-of-Hospital Cardiac Arrest Coverage Under Limited Resources. Resuscitation. 2018;125:83–89. 29414670 10.1016/j.resuscitation.2018.01.055 TravelTime: TravelTime API. 2024a. Reference Source TravelTime: traveltimeR: Interface to ‘Travel Time’ API. 2024b. Reference Source Etten J van : R Package gdistance: Distances and Routes on Geographical Grids. J. Stat. Softw. 2017;76(13):1–21. 10.18637/jss.v076.i13 Bergmann J von : rtraveltime. 2024. Reference Source Weiss DJ Andy Nelson HS Gibson WT : A Global Map of Travel Time to Cities to Assess Inequalities in Accessibility in 2015. Nature. 2018;553(7688):333–336. 29320477 10.1038/nature25181 Weiss DJ Andy Nelson CA Vargas-Ruiz K : Global Maps of Travel Time to Healthcare Facilities. Nat. Med. 2020;26(12):1835–1838. 32989313 10.1038/s41591-020-1059-1 Zahavi Y : Traveltime Budgets and Mobility in Urban Areas. United States: Federal Highway Administration;1974. Zhao J Elmore AJ Lee JSH : Replanting and Yield Increase Strategies for Alleviating the Potential Decline in Palm Oil Production in Indonesia. Agric. Syst. 2023;210:103714. 10.1016/j.agsy.2023.103714

Collectively >1600 citations per Google Scholar at the 28th of January 2025.

 10.21956/verixiv.1007.r1274 Reviewer response for version 1 Shi Hui 1 Referee University of California Santa Barbara, Santa Barbara, California, USA

Competing interests: No competing interests were disclosed.

 12 7 2025 Copyright: © 2025 Shi H 2025 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve-with-reservations

It's great to see the development of a package for estimating travel time. However, several areas could be improved to enhance its functionality and flexibility:

Input/Output Description: Please include a table that clearly outlines the input parameters and expected outputs. This will help users better understand how to interact with the package.

Walking Time Metrics: Currently, the package only provides the minimum walking time. For a more comprehensive comparison, it would be beneficial to also include walking times to the nearest few stations and the average walking time. Ideally, users should be able to specify the number of nearest stations they want to include.

Adjustable Resolution: Consider allowing users to set the spatial resolution themselves. This would make the package more applicable to larger areas, where a lower resolution may be preferred for performance or scalability.

 In short, I would love to see a more flexible and user-configurable package.

Are the conclusions about the tool and its performance adequately supported by the findings presented in the article?

Yes

Is the rationale for developing the new software tool clearly explained?

Partly

Is the description of the software tool technically sound?

Partly

Are sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others?

Partly

Is sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool?

Partly

Reviewer Expertise:

Spatial analysis and transportation

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

 10.21956/verixiv.1007.r1283 Reviewer response for version 1 Macharia Peter M 1 2 Referee https://orcid.org/0000-0003-3410-1881 Department of Public Health, Institute of Tropical Medicine, Antwerp, Belgium Population & Health Impact Surveillance Group, KEMRI-Wellcome Trust Research Programme Nairobi (Ringgold ID: 543945), Nairobi, Kenya

Competing interests: No competing interests were disclosed.

 8 7 2025 Copyright: © 2025 Macharia PM 2025 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve-with-reservations

This is a useful software package and timely for many accessibility applications to different services such as healthcare, education, and banking services. I have several points to note.

In paragraph three, there is need to further contextualise other packages and tools related to traveltime developed here, for example, gdistance, AccessMod, other APIs similar to TravelTime API

While a reference has been provided, it would be useful to have a paragraph that summarizes the data and process behind the creation of the friction surface, as it’s a major input in this package. This may include the characteristics of the data used and the approach.

A major reservation is the use of pre-existing friction surfaces. These friction surfaces were developed for global use; however, when used in a local or even country context, several key parameters determine the magnitude of travel time change. These include travel speeds; data on roads/pathways; combinations of modes of transport and sometimes differing land use; and also the output resolution.

While one solution offered here is an option for users to provide their own friction surface, If this package added a functionality such that the users can create a friction surface by controlling key input variables, it will not only ensure  each output is context-specific but all outputs will be updated or based on users’ time frame of interest

Are the conclusions about the tool and its performance adequately supported by the findings presented in the article?

Yes

Is the rationale for developing the new software tool clearly explained?

Yes

Is the description of the software tool technically sound?

Yes

Are sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others?

Yes

Is sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool?

Yes

Reviewer Expertise:

I am a spatial epidemiologist working at the interface of geospatial methods and their applications to global health challenges in SSA.   I am particularly interested in geographic access modelling and disease mapping.

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

 10.21956/verixiv.1007.r1195 Reviewer response for version 1 Sumner Michael D 1 Referee Australian Antarctic Division, Kingston, Tasmania, Australia

Competing interests: No competing interests were disclosed.

 17 6 2025 Copyright: © 2025 Sumner MD 2025 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve

This is an excellent pithy and practical paper. The authors present a toolkit for determining travel time across a landscape, which is a compelling problem in human health, transport logistics, and ecology to name a few. The toolkit uses modern programming practices, well-established foundations for calculating and using travel time, explaining how it works, and couches the approach within the potentially confusing ecosystem of software that already exists.

The work is justified in terms of its need in the software niche, its accessibility for immediate problem-solving, and the transparency and explanations of how it works. The software behind the paper and the results within are accessible on public websites, with all of its dependencies available on the public CRAN repository. The package passes a gamut of tests and documentation building that run any time a change is made. Uses can report problems and contribute software changes via an accessible mechanism.

I've installed the software, checked and built it on various systems, and gave a deep read of the software itself and how it works and I can say that the package itself is also concise, does not have any glaring technical debt or complexity problems. The software navigates the available ecosystem of R software by utilizing existing packages well and presenting a consistent user interface that can work with data obtained by the software, or by including extra local data by users themselves. 

This is a good piece of work and should be accepted. No changes needed.

Are the conclusions about the tool and its performance adequately supported by the findings presented in the article?

Yes

Is the rationale for developing the new software tool clearly explained?

Yes

Is the description of the software tool technically sound?

Yes

Are sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others?

Yes

Is sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool?

Yes

Reviewer Expertise:

I'm a geospatial expert with R, I have deep experience with animal tracking applications and good knowledge of how software like this works through ecosystem research.

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

Ryan Gerard The Kids Research Institue Australia, Nedlands, WA, Australia

Competing interests: No competing interests were disclosed.

 17 6 2025

Thank you Michael for this thorough consideration of our work and thoughtful comments.  Best wishes,  Gerry Ryan