The most important issues facing transportation sustainability and the development of future transportation systems are the crucial oil-supply issues we face today on a global scale and the need to reduce greenhouse gas emissions. The automobile industry will be called upon to respond to these issues. This requires an approach that emphasizes high efficiency and relatively low cost energy, and advocates electricity as a significant part of the solution to the transportation issues that the auto industry faces. A pivotal player in this new scheme is the plug-in hybrid electric vehicle (PHEV; Figure 1), which can very efficiently utilize supply-limited renewable and conventional electricity via a rechargeable energy storage unit.

Figure 1: Plug-in Hybrid Electric Vehicle (PHEV).  Image source: http://www.afdc.energy.gov

Before such systems can be manufactured for safe public use, fundamental questions such as battery electrochemistry, optimal PHEV battery-size, weight, life-cycle, optimal state of charge and performance characteristics need to be answered. Optimal PHEV battery characteristics need to be determined by measuring driver demands and producing duty cycles and taking into account optimal opportunistic recharging locations and time periods. Vehicle simulation needs to be performed and criteria developed which will optimize the displacement of fossil fuels with renewable electricity without sacrificing vehicle performance characteristics. Energy losses resulting from ineffective use of battery storage and engine horsepower need to be minimized.

During the past two years, our funded research team (Dr. Danny Blair, University of Winnipeg; Dr. David Strong, Queen’s University; Dr. Shaahin Filizadeh, University of Manitoba; Dr. Linda Nazaar, University of Waterloo; Dr. Eric Bibeau, University of Manitoba) used a grant provided by the federal government's AUTO21 - A Network of Centres of Excellence - research program to analyze the energy issues noted above.  The primary objective of the University of Winnipeg's research during this period was to characterize driving behaviour and duty cycles of urban drivers as represented by Winnipeg drivers.  GPS units recording the velocity, acceleration, location, trip lengths, trip frequencies, parking habits, and idling habits for about 75 private vehicles were monitored, second-by-second, for as long as one year per vehicle.  These GPS units were provided with assistance from the University of Winnipeg’s Centre for Sustainable Transportation and Persentech, Inc..  Furthermore, with the use of road information data stored within the GPS units (Otto Driving Companions, developed and produced by Persentech, Inc.; Figure 2), the frequency that drivers exceed posted speed limits was evaluated, and the frequency of visits to urban destinations, especially parking lots was characterized.  A full year of data collection was a much longer interval than normally utilized in transportation planning research, and research that uses GPS-based systems to study vehicle patterns has normally been restricted to smaller sample sizes.

Figure 2: Otto Driving Companion GPS device, developed by Winnipeg-based PERSENTECH.  Image source: Ryan Smith, 2010.

The AUTO21 research grant has recently been extended for a further two years with a total of $58,000. Our goal for these next two years is to continue to characterize driving behavior and to assess the battery chemistry, performance and storage requirements of PHEV battery systems.  In particular, a new group of Manitoba drivers will be selected who live in rural communities and commute regularly to Winnipeg for work and monitored for as long as one full year with GPS devices.  In the end, these travel patterns as reflected in trip lengths and durations, types of roads chosen, acceleration and speed, and times spent at locations that may be suitable locations for plug-in recharging facilities, will be fully assessed and characterized in the form of a set of “typical” 24-hour light duty vehicle (LDV) duty-cycle data, and delivered to our research partners at the University of Manitoba and elsewhere (Dr. Eric Bibeau’s research team).  

Methodology:

This project proposes to collect vehicle data for a full year, a much longer interval than normally utilized in transportation planning research. Research that uses GPS-based systems to study vehicle patterns is normally restricted to small sample sizes. Data is often collected using one car and a representative sample of drivers for a specified driving course (Holmen 1998 and Ericsson 2000). Axhausen et al. (2002) presented a six week driving study using a travel diary and indicated that it was the first such study in recent years. Transport planning has traditionally focused on a representative day and peak hour. While multi-day surveys of travel behavior have recently become more frequent, they are normally at most three days long and are motivated by an attempt to capture most of the variance of the behavior, but not its temporal structure (Axhausen et al. 2003). Lee et al. (2001) used GPS to improve data quality, to collect additional information and to simplify data analysis. Wolf et al. (2001) noted that the advantages of using GPS data were that they were automatically collected without burdening the respondent, and data routes were recorded for all trips with accurate trip start and end times, as well as trip lengths. To date no large data-set of vehicle activity has been analyzed for the purpose of PHEV optimization studies.

We will put GPS units (built by Persentech, Inc., in Winnipeg) in a maximum of 100 private vehicles operated in rural communities outside of the City of Winnipeg and which frequently travel to the City of Winnipeg, for a maximum period of one full year.  The GPS units will record the location of the vehicles every second, allowing the speed, acceleration, location and trip-lengths of the vehicles to be recorded with very high precision.  The resulting data set will be linked to a geographic data base containing information about road types and the locations of common urban destinations. 

The volumes of geographic data will be summarized and aggregated to produce a profile of driving habits and duty cycles.  The aggregation of the data will prevent the identities of individual drivers from being extracted from the data.   Furthermore, geographic references (i.e., latitudes and longitudes) presented in reports, publications or shared datasets will be geographically transformed to mask the locations of residences or places of work (that might be determined by examining the most frequently visited parking locations).  In the end, we (and others who may use the data) are not interested in the specific geographic patterns of vehicle travel related to individuals, but the patterns of travel and duty cycles as reflected in trip lengths and durations, types of roads chosen, acceleration and speed, and times spent at locations that may be suitable locations for plug-in recharging facilities; and we are interested in the relationships between these patterns and basic household socioeconomic characteristics (see Appendix A).

Participation in the research will be voluntary.  Volunteers will be recruited by the researchers, through word-of-mouth and media attention to our research project; thus, the sampling methodology is “snowball”. There is no incentive for participation.  At most 100 volunteers (vehicles) will be utilized in the year of sampling.

Communication of Findings:

Once completed, each participant will receive a summary of our findings.  These findings can be obtained by mail or email, and will also be posted on the research website: auto21.uwinnipeg.ca.

References:

Axhausen K.W., et al. (2002), “Observing the rhythms of daily life: A 6-week travel diary. Transportation,” 29 (2), 95-124.

Axhausen K.W. et al. (2003). “Eight weeks of GPS tracers: approaches to enrich trip

information,” Annual Meeting of the Transportation Research Board.

Lee M.S., et al. (2001), “Evaluation of a shared-use electric vehicle program: integrating a web-based survey with in-vehicle tracking,” Institute of Transportation Studies, Center for Activity Systems Analysis, University of California, Irvine, Paper No. UCI-ITS-AS-WP-01-5.

Wolf J., et al. (2001), “Elimination of the travel diary: An experiment to derive trip purpose from GPS travel data,” Paper No. 01-3255. Annual Meeting of the Transportation Research Board, Washington, D.C.