Context and Location Awareness in Eco-Driving Recommendations

Author(s)

Campolina, Andre ; Boukerche, Azzedine ; Loureiro, Antonio A. F. ; IEEE

• Document type: Journal article
• Source: NOMS 2020 - PROCEEDINGS OF THE 2020 IEEE/IFIP NETWORK OPERATIONS AND MANAGEMENT SYMPOSIUM 2020: MANAGEMENT IN THE AGE OF SOFTWARIZATION AND ARTIFICIAL INTELLIGENCE; v. N/A, p. 6-pg., 2020-01-01.

Abstract

Eco-driving techniques are methods that drivers can take in order to improve their vehicles' fuel efficiency. One way of implementing such methods is to recommend changes in driving habits focusing on saving fuel. Recommendation systems in the context of efficient driving may take numerous data sources as input and serve multiple purposes. In this work, we develop a recommendation system that suggests changes in engine revolutions per minute (RPM) that reduce fuel consumption. We simulate the effects of this system using a vehicular sensor dataset that contains location, speed, RPM, and consumption. We also investigate the effects of including location data in the recommendations as a way to leverage local habits and behaviors. Both methods reduced fuel consumption in all recorded trips. Moreover, using only local data yielded a mean fuel reduction of 43%, whereas using the entire dataset reduced the fuel consumption in 56% on average. Upon analyzing the resulting changes, we noted that such a difference in fuel consumption is due to the local recommendations not having access to global optimal data points and accounting for local behavior that is affected by aspects such as terrain. (AU)

• FAPESP's process: 18/23064-8 - Mobility in urban computing: characterization, modeling and applications (MOBILIS)
• Grantee: Antonio Alfredo Ferreira Loureiro
• Support Opportunities: Research Projects - Thematic Grants
• FAPESP's process: 15/24494-8 - Communications and processing of big data in cloud and fog computing
• Grantee: Nelson Luis Saldanha da Fonseca
• Support Opportunities: Research Projects - Thematic Grants