Applicability of Long Short-Term Memory Traffic Volume Imputation Model to Drive Connected Corridor Simulation




smart cities, connected corridor, LSTM, real-time simulation, traffic volume imputation


For effective implementation of connected corridor applications, it is imperative to study the characteristics of the high-resolution connected corridor data streams leveraged in smart city applications. In a previous effort, a smart city application – real-time corridor data-driven traffic simulation model, i.e. Digital Twin – is developed. Investigation of the corridor volume data streams revealed the presence of data gaps. To address the data gaps, deep Long Short-Term Memory Recurrent Neural Network univariate and multivariate volume imputation models are developed. In this paper, the impact of the developed model imputations on the digital twin generated travel times are investigated. Simulation runs are conducted for two scenarios – typical and atypical (holiday) traffic, for three volume input cases: base data (original volumes), univariate model imputations, and multivariate model imputations. Results indicated that: 1) the travel times generated using multivariate imputations are closer to the travel times generated using base data, 2) the impact of imputations on travel times is more focused on congested routes, and 3) the impact on travel time is minimal despite input volume overestimation on routes that have capacity to accommodate higher volumes. These finding demonstrate the need to prioritize data streams based on the given application and underlying corridor conditions.

Author Biographies

Dr. Abhilasha Saroj, Georgia Institute of Technology

Abhilasha Saroj received the M.S. degree and the Ph.D. degree in Civil Engineering from the Georgia Institute of Technology, Atlanta, USA, in August 2016 and August 2020, respectively. She is currently a Postdoctoral Fellow at School of Civil and Environmental Engineering, Georgia Institute of Technology.

Dr. Angshuman Guin, Georgia Institute of Technology

Angshuman Guin received his Ph.D. degree in Civil Engineering from the Georgia Institute of Technology in 2004. He is currently a senior research engineer at the Georgia Institute of Technology, Atlanta, Georgia. 

Dr. Michael Hunter, Georgia Institute of Technology

Michael Hunter received his Ph.D. in Civil Engineering from The University of Texas at Austin in 2003 and then joined the faculty of the Georgia Institute of Technology School of Civil and Environmental Engineering. Dr. Hunter’s research is unified by its integration of computer, communication, and sensor
technologies into the transportation infrastructure. He seeks to address current “real world” challenges by integrating emerging technologies and improved transportation data collection with enhanced operational analysis and modeling. Dr. Hunter has recently been leading the effort to build a real-time Digital Twin of the North Avenue Corridor in Atlanta, Ga.


USDOT. Connected Vehicle Pilot Deployment Program. n.d. [cited 2020 February 26]; Available from:

Frost, A. Singapore to develop first 5G C-V2X research testbed on NTU campus. 2019 October 22, 2019 [cited 2020 February 26 ]; Available from:

CenterForTransportationResearch. City of Austin Connected Corridors. 2018 [cited 2020 February 26]; Available from:

HighwaysEngland, Signs of the future: new technology testbed on the A2 and M2 in Kent. 2018, Highways England:

California CV Testbed. n.d. [cited 2020 February 26]; Available from:

Saroj, A., S. Roy, A. Guin, M. Hunter, and R.M. Fujimoto, Smart city real-time data-driven transportation simulation, in Proceedings of the 2018 Winter Simulation Conference. 2018, IEEE Press: Gothenburg, Sweden. p. 857–868.

Hunter, M., R. Guensler, A. Guin, A. Saroj, and S. Roy, Smart Cities Atlanta - North Avenue, in City of Atlanta Research Project. 2019: p. 82.

Saroj, A., Development of a real-time connected corridor data-driven digital twin and data imputation methods, in School of Civil and Environmental Engineering. 2020, Georgia Institute of Technology:

Kostadinov, S. How Recurrent Neural Networks work. 2017 December 2017 [cited 2020 February 22]; Available from:

See, A. Vanishing gradients and fancy RNNs. Natural Language Processing with Deep Learning 2019 2019 [cited 2020 July 14]; Available from:

Hochreiter, S. and J. Schmidhuber, Long Short-term Memory. Neural computation, 1997. 9: p. 1735-80.

Weber, N. Why LSTMs stop your gradients from vanishing: A view from the backwards pass. 2017 November 15 2017 [cited 2020 February 23]; Available from:

Yao, Y. and Z. Huang. Bi-directional LSTM Recurrent Neural Network for Chinese word segmentation. in Neural Information Processing. 2016. Cham: Springer International Publishing.