Skip to content

An open-source tool to identify active travel from hip-worn accelerometer, GPS and GIS data

Research output: Contribution to journalArticle

Standard

An open-source tool to identify active travel from hip-worn accelerometer, GPS and GIS data. / Procter, Duncan S.; Page, Angie S.; Cooper, Ashley R.; Nightingale, Claire M.; Ram, Bina; Rudnicka, Alicja R.; Whincup, Peter H.; Clary, Christelle; Lewis, Daniel; Cummins, Steven; Ellaway, Anne; Giles-Corti, Billie; Cook, Derek G.; Owen, Christopher G.

In: International Journal of Behavioral Nutrition and Physical Activity, Vol. 15, 91, 21.09.2018.

Research output: Contribution to journalArticle

Harvard

Procter, DS, Page, AS, Cooper, AR, Nightingale, CM, Ram, B, Rudnicka, AR, Whincup, PH, Clary, C, Lewis, D, Cummins, S, Ellaway, A, Giles-Corti, B, Cook, DG & Owen, CG 2018, 'An open-source tool to identify active travel from hip-worn accelerometer, GPS and GIS data', International Journal of Behavioral Nutrition and Physical Activity, vol. 15, 91. https://doi.org/10.1186/s12966-018-0724-y

APA

Procter, D. S., Page, A. S., Cooper, A. R., Nightingale, C. M., Ram, B., Rudnicka, A. R., ... Owen, C. G. (2018). An open-source tool to identify active travel from hip-worn accelerometer, GPS and GIS data. International Journal of Behavioral Nutrition and Physical Activity, 15, [91]. https://doi.org/10.1186/s12966-018-0724-y

Vancouver

Procter DS, Page AS, Cooper AR, Nightingale CM, Ram B, Rudnicka AR et al. An open-source tool to identify active travel from hip-worn accelerometer, GPS and GIS data. International Journal of Behavioral Nutrition and Physical Activity. 2018 Sep 21;15. 91. https://doi.org/10.1186/s12966-018-0724-y

Author

Procter, Duncan S. ; Page, Angie S. ; Cooper, Ashley R. ; Nightingale, Claire M. ; Ram, Bina ; Rudnicka, Alicja R. ; Whincup, Peter H. ; Clary, Christelle ; Lewis, Daniel ; Cummins, Steven ; Ellaway, Anne ; Giles-Corti, Billie ; Cook, Derek G. ; Owen, Christopher G. / An open-source tool to identify active travel from hip-worn accelerometer, GPS and GIS data. In: International Journal of Behavioral Nutrition and Physical Activity. 2018 ; Vol. 15.

Bibtex

@article{6b7713050cf74b82855314d3556e6c72,
title = "An open-source tool to identify active travel from hip-worn accelerometer, GPS and GIS data",
abstract = "Background: Increases in physical activity through active travel have the potential to have large beneficial effects on populations, through both better health outcomes and reduced motorized traffic. However accurately identifying travel mode in large datasets is problematic. Here we provide an open source tool to quantify time spent stationary and in four travel modes(walking, cycling, train, motorised vehicle) from accelerometer measured physical activity data, combined with GPS and GIS data. Methods: The Examining Neighbourhood Activities in Built Living Environments in London study evaluates the effect of the built environment on health behaviours, including physical activity. Participants wore accelerometers and GPS receivers on the hip for seven days. We time-matched accelerometer and GPS, and then extracted data from the commutes of 326 adult participants, using stated commute times and modes, which were manually checked to confirm stated travel mode. This yielded examples of five travel modes: walking, cycling, motorised vehicle, train and stationary. We used this example data to train a gradient boosted tree, a form of supervised machine learning algorithm, on each data point (131537 points), rather than on journeys. Accuracy during training was assessed using five-fold cross-validation. We also manually identified the travel behaviour of both 21 participants from ENABLE London (402749 points), and 10 participants from a separate study (STAMP-2, 210936 points), who were not included in the training data. We compared our predictions against this manual identification to further test accuracy and test generalisability. Results: Applying the algorithm, we correctly identified travel mode 97.3{\%} of the time in cross-validation (mean sensitivity 96.3{\%}, mean active travel sensitivity 94.6{\%}). We showed 96.0{\%} agreement between manual identification and prediction of 21 individuals’ travel modes (mean sensitivity 92.3{\%}, mean active travel sensitivity 84.9{\%}) and 96.5{\%} agreement between the STAMP-2 study and predictions (mean sensitivity 85.5{\%}, mean active travel sensitivity 78.9{\%}). Conclusion: We present a generalizable tool that identifies time spent stationary and time spent walking with very high precision, time spent in trains or vehicles with good precision, and time spent cycling with moderate precisionIn studies where both accelerometer and GPS data are available this tool complements analyses of physical activity, showing whether differences in PA may be explained by differences in travel mode. All code necessary to replicate, fit and predict to other datasets is provided to facilitate use by other researchers.",
keywords = "Accelerometer, Active travel, GPS, Gradient boosting, Machine learning, Physical activity, Travel mode, Xgboost",
author = "Procter, {Duncan S.} and Page, {Angie S.} and Cooper, {Ashley R.} and Nightingale, {Claire M.} and Bina Ram and Rudnicka, {Alicja R.} and Whincup, {Peter H.} and Christelle Clary and Daniel Lewis and Steven Cummins and Anne Ellaway and Billie Giles-Corti and Cook, {Derek G.} and Owen, {Christopher G.}",
year = "2018",
month = "9",
day = "21",
doi = "10.1186/s12966-018-0724-y",
language = "English",
volume = "15",
journal = "International Journal of Behavioral Nutrition and Physical Activity",
issn = "1479-5868",
publisher = "BioMed Central",

}

RIS - suitable for import to EndNote

TY - JOUR

T1 - An open-source tool to identify active travel from hip-worn accelerometer, GPS and GIS data

AU - Procter, Duncan S.

AU - Page, Angie S.

AU - Cooper, Ashley R.

AU - Nightingale, Claire M.

AU - Ram, Bina

AU - Rudnicka, Alicja R.

AU - Whincup, Peter H.

AU - Clary, Christelle

AU - Lewis, Daniel

AU - Cummins, Steven

AU - Ellaway, Anne

AU - Giles-Corti, Billie

AU - Cook, Derek G.

AU - Owen, Christopher G.

PY - 2018/9/21

Y1 - 2018/9/21

N2 - Background: Increases in physical activity through active travel have the potential to have large beneficial effects on populations, through both better health outcomes and reduced motorized traffic. However accurately identifying travel mode in large datasets is problematic. Here we provide an open source tool to quantify time spent stationary and in four travel modes(walking, cycling, train, motorised vehicle) from accelerometer measured physical activity data, combined with GPS and GIS data. Methods: The Examining Neighbourhood Activities in Built Living Environments in London study evaluates the effect of the built environment on health behaviours, including physical activity. Participants wore accelerometers and GPS receivers on the hip for seven days. We time-matched accelerometer and GPS, and then extracted data from the commutes of 326 adult participants, using stated commute times and modes, which were manually checked to confirm stated travel mode. This yielded examples of five travel modes: walking, cycling, motorised vehicle, train and stationary. We used this example data to train a gradient boosted tree, a form of supervised machine learning algorithm, on each data point (131537 points), rather than on journeys. Accuracy during training was assessed using five-fold cross-validation. We also manually identified the travel behaviour of both 21 participants from ENABLE London (402749 points), and 10 participants from a separate study (STAMP-2, 210936 points), who were not included in the training data. We compared our predictions against this manual identification to further test accuracy and test generalisability. Results: Applying the algorithm, we correctly identified travel mode 97.3% of the time in cross-validation (mean sensitivity 96.3%, mean active travel sensitivity 94.6%). We showed 96.0% agreement between manual identification and prediction of 21 individuals’ travel modes (mean sensitivity 92.3%, mean active travel sensitivity 84.9%) and 96.5% agreement between the STAMP-2 study and predictions (mean sensitivity 85.5%, mean active travel sensitivity 78.9%). Conclusion: We present a generalizable tool that identifies time spent stationary and time spent walking with very high precision, time spent in trains or vehicles with good precision, and time spent cycling with moderate precisionIn studies where both accelerometer and GPS data are available this tool complements analyses of physical activity, showing whether differences in PA may be explained by differences in travel mode. All code necessary to replicate, fit and predict to other datasets is provided to facilitate use by other researchers.

AB - Background: Increases in physical activity through active travel have the potential to have large beneficial effects on populations, through both better health outcomes and reduced motorized traffic. However accurately identifying travel mode in large datasets is problematic. Here we provide an open source tool to quantify time spent stationary and in four travel modes(walking, cycling, train, motorised vehicle) from accelerometer measured physical activity data, combined with GPS and GIS data. Methods: The Examining Neighbourhood Activities in Built Living Environments in London study evaluates the effect of the built environment on health behaviours, including physical activity. Participants wore accelerometers and GPS receivers on the hip for seven days. We time-matched accelerometer and GPS, and then extracted data from the commutes of 326 adult participants, using stated commute times and modes, which were manually checked to confirm stated travel mode. This yielded examples of five travel modes: walking, cycling, motorised vehicle, train and stationary. We used this example data to train a gradient boosted tree, a form of supervised machine learning algorithm, on each data point (131537 points), rather than on journeys. Accuracy during training was assessed using five-fold cross-validation. We also manually identified the travel behaviour of both 21 participants from ENABLE London (402749 points), and 10 participants from a separate study (STAMP-2, 210936 points), who were not included in the training data. We compared our predictions against this manual identification to further test accuracy and test generalisability. Results: Applying the algorithm, we correctly identified travel mode 97.3% of the time in cross-validation (mean sensitivity 96.3%, mean active travel sensitivity 94.6%). We showed 96.0% agreement between manual identification and prediction of 21 individuals’ travel modes (mean sensitivity 92.3%, mean active travel sensitivity 84.9%) and 96.5% agreement between the STAMP-2 study and predictions (mean sensitivity 85.5%, mean active travel sensitivity 78.9%). Conclusion: We present a generalizable tool that identifies time spent stationary and time spent walking with very high precision, time spent in trains or vehicles with good precision, and time spent cycling with moderate precisionIn studies where both accelerometer and GPS data are available this tool complements analyses of physical activity, showing whether differences in PA may be explained by differences in travel mode. All code necessary to replicate, fit and predict to other datasets is provided to facilitate use by other researchers.

KW - Accelerometer

KW - Active travel

KW - GPS

KW - Gradient boosting

KW - Machine learning

KW - Physical activity

KW - Travel mode

KW - Xgboost

UR - http://www.scopus.com/inward/record.url?scp=85053660690&partnerID=8YFLogxK

U2 - 10.1186/s12966-018-0724-y

DO - 10.1186/s12966-018-0724-y

M3 - Article

VL - 15

JO - International Journal of Behavioral Nutrition and Physical Activity

JF - International Journal of Behavioral Nutrition and Physical Activity

SN - 1479-5868

M1 - 91

ER -