Passenger car fuel economy: Insights from big data analytics

Year : 2026 | Volume : 13 | Issue : 01 | Page : 8 19
    By

    Arumugam P,

  • Karthikeyan Subramanian,

  • Rajavel Rangasamy,

  1. Research Scholar, Department of Marine Engineering, AMET University, 603112, Tamilnadu, Chennai, India
  2. Assistant General Manager, Fuel Cell Product Development, Ashokleyland Technical Centre, Chennai 600103, Tamilnadu, India
  3. HOD, Department Marine Engineering, Faculty of Engineering, AMET University, Chennai 603112, Tamilnadu, India

Abstract

The increasing availability of real-world vehicle telematics through On-Board Diagnostics II (OBD-II) systems has enabled data-driven evaluation of passenger car fuel economy beyond conventional laboratory-based test cycles. While standardized certification procedures ensure repeatability, they often fail to capture the influence of real-world traffic conditions, driver behaviour, and transient vehicle operation. This study presents a structured Big Data Analytics (BDA) approach for analysing high-frequency OBD-II data collected from a gasoline passenger vehicle operated under urban, highway, and mixed driving conditions representative of Indian roads.

The dataset consists of engine speed, vehicle speed, throttle position, engine load, longitudinal acceleration, idling characteristics, intake air temperature, coolant temperature, and fuel-related parameters recorded at one-second resolution during continuous on-road operation. A systematic data processing methodology was adopted, including data sanity checks, filtering of invalid records, normalization, and selection of physically meaningful parameters to ensure analytical reliability. Descriptive statistical analysis and correlation-based evaluation were applied to examine interactions between key operating variables and fuel economy behaviour.

To further understand the influence of driving behaviour, clustering-based analysis was employed to identify dominant operating patterns directly from the data without reliance on predefined driving labels. In addition, regression-based sensitivity analysis was used to quantify the relative impact of key parameters on real-world fuel economy under varying driving conditions. The results indicate that transient acceleration events, engine load variation, idling duration, and speed instability play a significant role in fuel economy degradation, particularly under congested and mixed driving conditions.

Overall, the study demonstrates that real-world fuel economy is governed by the combined interaction of multiple vehicle operating and driver-related factors rather than isolated parameters. The proposed low-cost OBD-II data analytics framework provides practical engineering insights into passenger car fuel economy under actual driving conditions and supports more realistic assessment of fuel efficiency for applications related to energy consumption, emissions reduction, and sustainable mobility planning.

Keywords: Fuel Economy; OBD-II Data; Big Data Analytics; Driving Behaviour; Real-World Vehicle Operation; Passenger Car

[This article belongs to Journal of Automobile Engineering and Applications ]

How to cite this article:
Arumugam P, Karthikeyan Subramanian, Rajavel Rangasamy. Passenger car fuel economy: Insights from big data analytics. Journal of Automobile Engineering and Applications. 2026; 13(01):8-19.
How to cite this URL:
Arumugam P, Karthikeyan Subramanian, Rajavel Rangasamy. Passenger car fuel economy: Insights from big data analytics. Journal of Automobile Engineering and Applications. 2026; 13(01):8-19. Available from: https://journals.stmjournals.com/joaea/article=2026/view=236327


References

[1]   Kumar R, Jain A. Driving behavior analysis and classification by vehicle OBD data using machine learning. J Supercomputing. 2023;79(3):3367–3394.

[2]  Meseguer JE, Calafate CT, Cano JC, Manzoni P. DrivingStyles: a mobile platform for driving styles and fuel consumption characterization. arXiv preprint arXiv:1611.09065. 2016.

[3]    Kumar R, Singh S. Design and development of OBD-II compliant driver information system. Indian J Sci Technol. 2020;13(4):412–418.

[4]  Khan MAA, Islam MR, Hossain MS. A machine learning approach for driver identification based on CAN-BUS sensor data. arXiv preprint arXiv:2207.10807. 2022.

[5] Dong W, Li L, Wang Y, Wang K. Characterizing driving styles with deep learning. arXiv preprint arXiv:1607.03611. 2016.

[6]    Narayanan SN, Mittal S, Joshi A. Using data analytics to detect anomalous states in vehicles. arXiv preprint arXiv:1512.08048. 2015.

[7]   Verma A. OBD-II based intelligent vehicular diagnostic system using IoT. Int J Sci Res Comput Sci Eng. 2019;7(1):23–28.

[8]  Bhalerao S, Dhamankar A, Phadke A. OBD-II based intelligent vehicular diagnostic system using IoT. Int J Sci Res Comput Sci Eng. 2022;10(1).

[9]    Ramai C, Kechadi T, Young JS. A framework for building low-cost OBD-II data-logging systems. Vehicles. 2022;4(4):1209–1224.

[10] Patil P. Study of machine learning based applications on OBD-II port by using internet of vehicles. Int J Res Appl Sci Eng Technol. 2019;7(5):1498–1500.

[11] Ghosh A, Basak R. Data analytics in intelligent transportation systems. SAE Tech Pap. 2016;2016-01-1423.

[12]  Sharma S, Mehta R. A survey on vehicle diagnostic with OBD-II telematics. Int J Res Anal Rev. 2019;6(2):102–108.

[13]  Xu W, Zhou H, Zhang Z. Internet of vehicles in the big data era: a survey. IEEE/CAA J Autom Sin. 2018;5(1):19–35.

[14] Michailidis ET, Panagiotopoulou A, Papadakis A. A Review of OBD-II-Based Machine Learning Applications for Sustainable, Efficient, Secure, and Safe Vehicle Driving. Sensors. 2025 Jun 29;25(13):4057.

[15]  Lv Y, Duan Y, Kang W, Li Z, Wang F. Traffic flow prediction with big data: a deep learning approach. IEEE Trans Intell Transp Syst. 2015;16(2):865–873.

[16]  Girma A, Yan X, Homaifar A. Driver identification based on vehicle telematics data using LSTM. arXiv preprint arXiv:1911.08030. 2019.

[17] Michailidis K, Bose A, Nayak R. A comprehensive survey of machine learning models applied to OBD-II vehicle data. J Intell Veh Syst. 2025;12(1):11–38.

[18] Rykała A, Krawczyk M, Nowak T. Experimental comparison of ML methods for real-time fuel prediction using OBD-II. IEEE Trans Veh Technol. 2023;72(4):1552–1560.

[19]  Manjunath S, Kumar V. Predicting fuel economy using support vector regression on real driving data. Int J Automot AI. 2024;11(1):44–52.

[20]  Patel R, Natarajan L, Lee K. Hybrid machine learning model for vehicle gear prediction and fuel estimation using OBD-II and GPS. Sensors Veh J. 2025;9(3):92–101.

[21]  Heywood JB. Internal combustion engine fundamentals. 2nd ed. New York: McGraw-Hill; 2018.

[22]  Rakha H, Ahn K. Vehicle dynamics and fuel consumption under real-world driving conditions. Transp Res Part D. 2004;9(6):379–395.

[23]  Barth M, Boriboonsomsin K. Real-world modal fuel consumption models. Transp Res Rec. 2008;(2058):163–171.

[24]  Ericsson E. Independent driving pattern factors and their influence on fuel-use. Transp Res Part D. 2001;6(5):325–345.

[25]  Bishop J. The effect of driving style on fuel consumption. Proc IMechE Part D. 2012;226(1):78–92.

[26]  Rakha H, Ahn K, Trani A. Development of VT-Micro model for estimating hot stabilized light duty vehicle and truck emissions. Transportation Research Part D: Transport and Environment. 2004 Jan 1;9(1):49-74.

[27]  Fontaras G, Samaras Z. Fuel consumption modeling of passenger cars. Energy Policy. 2010;38(10):5668–5678.

[28]  Yang L, Zhang S, Wu Y, Chen Q, Niu T, Huang X, Zhang S, Zhang L, Zhou Y, Hao J. Evaluating real-world CO2 and NOX emissions for public transit buses using a remote wireless on-board diagnostic (OBD) approach. Environmental pollution. 2016 Nov 1;218:453-62.

[29]  Guo J, Liu Y, Chen H. Real-time fuel rate estimation using CAN-Bus data. IEEE Trans Veh Technol. 2018;67(10):9347–9358.

[30]  De Vlieger I. On-road emission and fuel consumption measurement. Atmos Environ. 2000;34(27):4621–4628.

[31] Ahn K, Rakha H. Fuel consumption modeling based on instantaneous speed and acceleration. Transp Res Part D. 2002;7(6):479–494.

[32]   Jiménez-Palacios J. Understanding engine efficiency maps under transient operation [PhD thesis]. Cambridge (MA): Massachusetts Institute of Technology; 1999.

[33]    Sun C, Barth M, Boriboonsomsin K. Impact of idling and stop-and-go traffic on fuel consumption. Appl Energy. 2015;151:153–165.

[34] Demir E, Bektaş T, Laporte G. Fuel consumption modeling with speed–load relationships. Eur J Oper Res. 2014;238(1):206–216.

[35]   Lee J, Park S, Kim J. Analysis of throttle and engine load influence on fuel consumption. SAE Tech Pap. 2013.

[36]   Vancoillie J, Verhelst S, Demuynck J. Gear selection and fuel economy under real driving conditions. Energy. 2013;53:145–156.

[37]   Wang J, Zhang L, Li Y. Driver behavior recognition using vehicle signals. IEEE Trans Intell Transp Syst. 2016;17(6):1612–1622.

[38]   Wu Y, Zhang J, Liu X. Driving behavior and fuel efficiency analysis using OBD data. Energy. 2017;115:1058–1067.

[30]   Park J, Lee S, Kim H. CAN-Bus-based energy consumption estimation for vehicles. IEEE Sens J. 2020;20(15):8523–8532.

[40]    Bishop R. Intelligent vehicle R&D: A review and contrast of programs worldwide and emerging trends. InAnnales des télécommunications 2005 Apr (Vol. 60, No. 3, pp. 228-263). Paris: Springer-Verlag.

Regular Issue Subscription Original Research
Volume 13
Issue 01
Received 18/12/2025
Accepted 21/01/2026
Published 10/02/2026
Publication Time 54 Days
1

Login


My IP

PlumX Metrics