A methodology to evaluate driving efficiency for professional drivers based on a maturity model

• Published in: Transportation research. Part C, Emerging technologies Emerging technologies, 2017-12, Vol.85, p.148-167
• Main Authors: Pozueco, Laura, Pañeda, Xabiel G., Tuero, Alejandro G., Díaz, Gabriel, García, Roberto, Melendi, David, Pañeda, Alejandro G., Sánchez, José A.
• Format: Article
• Language: English
• Subjects: Adaptive-learning; Dashboard; Efficient driving; Learning evaluation methodology; On-the-job training
• ISSN: 0968-090X; 1879-2359
• DOI: 10.1016/j.trc.2017.09.017

[Display omitted] •Methodology to evaluate driving efficiency based on driving patterns.•Initial stage with an early-classification of driving behavior to adapt the learning.•Periodic evaluation phase with maturity models to allow a continuous improvement.•Dashboard to support the evaluation methodology and perform detailed analysis.•Feedback reports with personalized recommendations to improve driving efficiency. Over the last decade, transport companies have tried to reduce fuel consumption using efficient driving programs. In them, motorists have to apply different specific techniques while driving. Thus, to succeed in this learning process there are two key elements: the knowledge of efficient driving techniques and the drivers’ motivation. The latter is a human factor which companies usually bring about by using reward systems. In this case, having a fair evaluation mechanism is the keystone to determine goal fulfilment. This paper presents a complete methodology to evaluate driving efficiency of drivers in professional fleets. The evaluation methodology is based on a continuous process which determines the maturity of the motorist in different aspects, such as the efficiency during the start of the vehicle movement, during motion or in stop events. In addition, the evaluation methodology includes an early-classification method to establish the initial efficiency level of the individual drivers which permits an adaptation of the learning process from the beginning. A dashboard has also been developed to support the evaluation methodology. 880 professional drivers have been evaluated with this methodology. Results show that the evaluation methodology identifies drivers’ weaknesses, to be improved in successive iterations of the learning process.