Estimation of Missing or Incomplete Data in Road Performance Measurement Systems

Authors: Kristjan Kuhi, Kati K. Kaare, Ott Koppel

Abstract:

Modern management in most fields is performance based; both planning and implementation of maintenance and operational activities are driven by appropriately defined performance indicators. Continuous real-time data collection for management is becoming feasible due to technological advancements. Outdated and insufficient input data may result in incorrect decisions. When using deterministic models the uncertainty of the object state is not visible thus applying the deterministic models are more likely to give false diagnosis. Constructing structured probabilistic models of the performance indicators taking into consideration the surrounding indicator environment enables to estimate the trustworthiness of the indicator values. It also assists to fill gaps in data to improve the quality of the performance analysis and management decisions. In this paper authors discuss the application of probabilistic graphical models in the road performance measurement and propose a high-level conceptual model that enables analyzing and predicting more precisely future pavement deterioration based on road utilization.

Keywords: Probabilistic graphical models, performance indicators, road performance management, data collection

Digital Object Identifier (DOI): doi.org/10.5281/zenodo.1088120

References:

[1] A. Neely, M. Gregory, and K. Platts, “Performance Measurement System Design. A Literature Review and Research Agenda,” Int. J. Operations & Production Management, Vol. 15, no. 4, pp. 80-116, 1995.
[2] T. Wegelius-Lehtonen, “Performance measurement in construction logistics,” Int. J. Production Economics, no. 69, pp. 107-116, 2001.
[3] S. Tangen, “Performance management: from philosophy to practice,” Int. J. Productivity and Performance Management, Vol. 53, No. 8, pp. 726-737, 2004.
[4] Performance Measures for Road Networks: A Survey of Canadian Use. Transportation Association of Canada, 2006. 67 p.
[5] J. Litzka, B. Leben, F. La Torre, A. Weninger-Vycudil, M. de Lurdes Antunes, D. Kokot, G. Mladenovic, S. Brittain, and H. Viner, The way forward for pavement performance indicators across Europe. Vienna: Austrian Transportation Research Association, 2008. 68 p.
[6] M. A. Ismail, R. Sadiq, H. R. Soleymani, and S. Tesfamariam, “Developing aroad performance index using a Bayesian belief network model,” J. Franklin Institute, no. 348, pp. 2539-2555, 2011.
[7] Performance Indicators for the Road Sector. Summary of the field tests. Paris: OECD Publications, 2001. 85 p.
[8] Z. J. Radnor, and D. Barnes, “Historical analysis of performance measurement and management in operations management,” Int. J. Productivity and Performance Management, no. 56(5/6), pp. 384-396, 2007.
[9] R. Haas, G. Felio, Z. Lounis, and L. C. Falls, “Measurable Performance Indicators for Roads: Canadian and International Practice,” in Proc. Annual Conf. Transportation Association of Canada, Vancouver, 2009.
[10] D. Osborne, and T. Gaebler, Reinventing Government. Boston: Addison- Wesley, 1992. 427 p.
[11] K. Kõrbe Kaare, Performance Measurement of a Road Network: A Conceptual and Technological Approach for Estonia. Tallinn: TUT Press, 2013. 162 p.
[12] K. K. Kaare, and O. Koppel, “Performance measurement data as an input in transportation policy,“ in Conf. papers of the XXVIII Int. Baltic Road Conf., Vilnius, 2013.
[13] J. Weimer, Y. Xu, C. Fischione, K. H. Johansson, P. Ljungberg, C. Donovan, A. Sutor, and L. E. Fahlén, “A Virtual Laboratory for Micro-Grid Information and Communication Infrastructures,” in 3rdIEEE PES Innovative Smart Grid Technologies Europe (ISGT Europe), Berlin, 2012.
[14] P. C. Zikopoulos, C. Eaton, D. deRoos, T. Deutsch, and G. Lapis, Understanding Big Data. Analytics for Enterprise Class Hadoop and Streaming Data. McGraw-Hill, 2012. 141 p.
[15] J. Dean, and S. Ghemawat, “MapReduce: Simplified Data Processing on Large Clusters,” Communications of the ACM, Vol. 51, no. 1, pp. 107- 113, 2008.
[16] K. Kõrbe Kaare, K. Kuhi, and O. Koppel, “Tire and pavement wear interaction monitoring for road performance indicators,” Estonian J. Eng., no. 18(4), pp. 324-335, 2012.
[17] K. Kõrbe Kaare, K. Kuhi, and O. Koppel, “Developing Road Performance Measurement System with Evaluation Instrument,” World Academy of Science, Engineering and Technology, no. 72, pp. 90-96, 2012.
[18] Pavement management and performance. California Department of Transportation, 2012.http://www.dot.ca.gov/hq/maint/Pavement/Offices/Pavement_Man agement/ (accessed 15.07.2013).
[19] M. de Lurdes Antunes (Ed.), Framework for implementation of Environment Key Performance Indicators. EVITA (Environmental Indicators for the Total Road Infrastructure Assets) Deliverable D4.1. InstitutFrançais des Sciences et des Technologies desTransports, de l’Aménagement et des Réseaux and PMS-Consult, 2011. 19 p.
[20] Annual Report 2012. Tallinn: Estonian Road Administration, 2013. 74 p.
[21] P. H. Ibargüengoytia, U. A. Garcıa, J. Herrera-Vega, P. Hernandez-Leal, E. F. Morales, L. E. Sucar, and F. Orihuela-Espina, “On the Estimation of Missing Data in Incomplete Databases: Autoregressive Bayesian Networks,” in ICONS 2013, Seville, 2013, pp. 111-116.
[22] D. Smith, G. Timms, P. De Souza, and C.D'Este, “A Bayesian Framework for the Automated Online Assessment of Sensor Data Quality,” Sensors (Basel), 12(7), pp. 9476–9501, 2012.
[23] T. Jairus, “Estonian Road Information System Tark tee (Smart road), in Conf. papers of the XXVIII Int. Baltic Road Conf., Vilnius, 2013.