Verification of the HDM-4 fuel consumption model using a Big data approach: A UK case study.

Highlights • The paper presents a verification of HDM-4 fuel consumption estimates by comparing these to real world measurements from a large dataset from a truck telematic database. • Results of the study show that the latest calibration of HDM-4 performed in UK is not able to estimate the fuel consumption of the considered fleet of trucks reliably. • The authors suggest that, in order to use HDM-4 at a strategic level in road asset management, a full recalibration of the model should be performed in UK. • Results show also that analysis of the 'big data' available in the telematic database of fleet managers can help in updating HDM-4 without performing new experiments and can represent a valid alternative to the traditional approach. Abstract This paper presents an assessment of the accuracy of the HDM-4 fuel consumption model calibrated for the United Kingdom and evaluates the need for further calibration of the model. The study focuses on HGVs and compares estimates made by HDM-4 to measurements from a large fleet of vehicles driving on motorways in England. The data was obtained from the telematic database of truck fleet managers (SAE J1939) and includes three types of HGVs: light, medium and heavy trucks. Some 19,991 records from 1645 trucks are available in total. These represent records of trucks driving at constant speed along part of the M1 and the M18, two motorways in England. These conditions have been simulated in HDM-4 by computing fuel consumption for each truck type driving at a constant speed of 85 km/h on a flat and straight road segment in good condition. Estimates are compared to real measurements under two separate sets of assumptions. First, the HDM-4 model calibrated for the UK has been used. Then, the model was updated to take into account vehicle weight and frontal area specific to the considered vehicles. The paper shows that the current calibration of HDM-4 for the United Kingdom already requires recalibration. The quality of the model estimates can be improved significantly by updating vehicle weight and frontal area in HDM-4. The use of HGV fleet and network condition data as described in this paper provides an opportunity to verify HDM-4 continuously. [ABSTRACT FROM AUTHOR]