Evaluating the mileage and time efficiency of ridesourcing services: Austin, Texas case

Figures & data

Table 1. A summary of evidence from the literature on the efficiency of ridesourcing services

	Sample description
Research	Driver sample size	Data type	Data period	City/Region	Service provider	Research findings*	Comments
Cramer and Krueger (2016)	Randomlyselected 2,000 UberX drivers	Unknown	December 1 2014 to December 1 2015	Los Angeles New York Seattle	Uber	Mileage efficiency is 55.2% for Seattle and 64.2% for Los Angeles.	The results are directly provided by Uber and potential inclusion of personal travel because if the app is open during those periods.
			From July to October 2015 Average of January 9, April 11, and July 13 in 2015	San Francisco Boston		Time efficiency is 50.3% for Los Angles, 43.6% for Seattle, 51.2% for New York, 54.3% for San Francisco, 46.1% for Boston.
SFCTA,(2017)	Unknown	API Data	Mid-November to Mid-December 2016, excluding dates around the Thanksgiving 2016 holiday.	San Francisco	Uber, Lyft	Mileage efficiency is around 80% for intra-city trips.	Analysis based on approximate vehicle locations at five-second intervals, estimated passenger pick-up times, and active vehicle time on the app. Potential overestimation because empty distances from ride request to pick-up locations were taken as passenger ride.
RockyMountain Institute, (2018) (as cited in Said (2018))	Unknown	Unknown	November 1 2016 to October 31 2017	Chicago New York San Francisco	Lyft	The average distances of cruising, over-heading, and passenger rides are indicated for all three cities.	Given the average distance data of each segment, efficiency in Chicago, New York and San Francisco can be measured as 59.5%, 59.3%, and 67.2%, respectively. Potential inclusion of personal travel in cruising parts.
Henao and Marshall (2019)	1 driver in both services	416 recorded trips data	14 weeks during fall 2016	Denver	Uber, Lyft	Mileage efficiency is 59.2%, while excluding the commuting part it is measured as 70.3%.	Conservative strategy followed to minimize the deadheading (e.g. using two different services at the same time, not taking a passenger whose location is farther than 15 miles, choosing specific start locations, and preferring to work during peak demand times of the day).
Henao (2017)	Same as above					Time efficiency is 41.3%, excluding commuting at the end shift.
George and Zafar (2018)	Unknown	Unknown	October 2017	California	Uber, Lyft	Mileage efficiency is around 60%.	Potential underestimation because most of the drivers use more than one apps at the same time (e.g. trip miles traveled on one platform may be recorded as deadhead miles on another). The results are directly provided by TNC operators.
Komanduri et al. (2018)	Appr. 5,000 drivers	1,494,125 recorded trips data	June 4 2016 to April 13 2017	Austin, Texas	RideAustin	Mileage efficiency is 67%, while time efficiency is 50%, including commuting parts.	Conservative deadhead measurement since the straight-line distances between rides was considered.
Xue et al. (2018)	13,985 drivers	Daily orders data	From the August to December (the year is unknown)	Beijing	DidiChuXing	The average trip distance is about 8.5 km, and the over-heading distance is nearly 25% of the total trip distance.	Given theinformation, overestimated efficiency can be measured as 80% because of excluding cruising part of deadheading.
Wenzel et al. (2019)	Appr. 5,000 drivers	Recorded trips data	October 27 2016 to April 13 2017	Austin, Texas	RideAustin	Mileage efficiency is 55%, including commuting parts, while 64.5% for less than 60 minutes in between consecutive rides.	Applied the correction factor on straight-line distances between locations to represent the street network distance.
Tu et al. (2019)	144,867 drivers	GPStrajectory data	November 24 2015 to November 30 2015	Beijing	Didi ChuXing	The average daily total travel distance of drivers is 129.4 km and the average trip length is 9.7 km, while average daily travel count is 9.	Given the average total travel distance and passenger ride distance, efficiency is estimated at 67.5% – (9x9.7)/129.4.
City of Toronto (2019)	Unknown	Unknown	For March 2017 and September 2018	Toronto	Uber	55% of the total VMT was spent in-service.	The results are directly provided by Uber and details of findings are not known.

it should be noted that some studies may include the pooled ridesourcing trip (e.g., UberPool and LyftLine) (Henao 2017; Henao and Marshall 2019). Clearly mentions inclusion of pooled trips, others do not explicitly mention anything about pooling.

Figure 1. The total number of rides, active drivers, and the average number of rides per driver, by week of selected 200 drivers.

Figure 2. (a) Sample of RideAustin drivers’ working time flow, with empty trip segments shown in the dashed line. Travel segments of drivers in between two rides: (b) dispatched time after previous ride end; (c) dispatched time before previous ride end.

Figure 3. Components of two consecutive trips sample (representational, does not reflect the actual ride performed)

Figure 4. (a) The number of trips by the time of the day. (b) Average miles per trip by the time of day

Figure 5. Cumulative distribution of passenger rides and empty trips in the drivers’ shift according to cutoff time limits (for 200 busiest drivers)

Figure 6. Mileage efficiency by time cutoff points for breaks

Figure 7. Distribution of the mileage efficiency of drivers for 60 minutes time limit: (a) by density; (b) by average working time per shift; (c) by the average number of trips per shift; (d) by the differences between scenarios

Table 2. Mileage efficiency rate statistic according to the category of drivers

Driver category	Trips per shift (n)	Average trips per shift	Worst-case mileage efficiency (%)				Best-case mileage efficiency (%)
			Min.	Max.	Mean	S.D.	Min.	Max.	Mean	S.D.
1^st group	n ≤ 4.6	4.0	42.62	57.90	50.02	3.21	61.33	76.63	68.70	2.84
2^nd group	4.6 ≤ n < 6	5.4	44.21	56.73	50.83	3.26	60.66	73.60	67.10	2.26
3^rd group	6 ≤ n < 8.2	6.8	43.30	54.83	50.06	2.80	59.05	73.01	66.45	2.78
4^th group	8.2 ≤ n	9.7	44.65	57.69	49.92	2.94	60.74	71.72	65.11	2.59

Figure 8. Mileage efficiency: by day of the week (a) worst-case, (b) best-case scenario; (c) by the time of day. (d) Number of drivers worked according to day of week and time of day

Figure 9. Time efficiency rate (a) by the time cutoff points for break (b) distribution of 200 drivers for 60 minutes time limit (c) by the average number of trips per shift

Figure 10. Time efficiency (a) by day of the week (b) by the time of day

Cramer, J., and A. B. Krueger. 2016. “Disruptive Change in the Taxi Business: The Case of Uber.” American Economic Review 106: 177–182. doi:https://doi.org/10.1257/aer.p20161002.

 Google Scholar

San Francisco County Transportation Authority, 2017. “TNCs Today: A Profile of San Francisco Transportation Network Company Activity”. https://www.sfcta.org/projects/tncs-today

 Google Scholar

Said, C., 2018. “Lyft Trips in San Francisco More Efficient than Personal Cars, Study Finds”. https://www.sfchronicle.com/business/article/Lyft-trips-in-San-Francisco-more-efficient-than-12476962.php

 Google Scholar

Henao, A., and W. E. Marshall. 2019. The Impact of Ride-hailing on Vehicle Miles Traveled. Transportation 46, 2173–2194. doi:https://doi.org/10.1007/s11116-018-9923-2

 Google Scholar

Henao, A. 2017. Impacts of Ridesourcing-Lyft and Uber-on Transportation Including VMT, Mode Replacement, Parking, and Travel Behavior. University of Colorado at Denver. https://digital.auraria.edu/content/AA/00/00/60/55/00001/Henao_ucdenver_0765D_10823.pdf

 Google Scholar

George, S. R., and M. Zafar, 2018. “Electrifying the Ride-Sourcing Sector in California. California Public Utilities Commission”. https://www.cpuc.ca.gov/ppd_work

 Google Scholar

Komanduri, A., Z. Wafa, K. Proussaloglou, and S. Jacobs. 2018. “Assessing the Impact of App-Based Ride Share Systems in an Urban Context: Findings from Austin.” Transportation Research Record 2672: 34–46. doi:https://doi.org/10.1177/0361198118796025.

 Web of Science ®Google Scholar

Xue, M., B. Yu, Y. Du, B. Wang, B. Tang, and Y. M. Wei. 2018. “Possible Emission Reductions from Ride-Sourcing Travel in a Global Megacity: The Case of Beijing.” The Journal of Environment & Development 27: 156–185. doi:https://doi.org/10.1177/1070496518774102.

 Web of Science ®Google Scholar

Wenzel, T., C. Rames, E. Kontou, and A. Henao. 2019. “Travel and Energy Implications of Ridesourcing Service in Austin, Texas.” Transportation Research Part D: Transport and Environment 70: 18–34. doi:https://doi.org/10.1016/j.trd.2019.03.005.

 Web of Science ®Google Scholar

Tu, W., P. Santi, T. Zhao, X. He, Q. Li, L. Dong, T. J. Wallington, and C. Ratti. 2019. “Acceptability, Energy Consumption, and Costs of Electric Vehicle for Ride-hailing Drivers in Beijing.” Applied Energy 250: 147–160. doi:https://doi.org/10.1016/j.apenergy.2019.04.157.

 Web of Science ®Google Scholar

City of Toronto, 2019. “Research & Analysis: The Transportation Impacts of Vehicle-for-Hire in the City of Toronto”. https://www.toronto.ca/wp-content/uploads/2019/06/96c7-Report_v1.0_2019-06-21.pdf

 Google Scholar