*3.1. Data Input and Preprocessing*

For the case study, vehicle availabilities at home are computed based on the data sets of the Californian Household Travel Survey (US CHTS) and the German Mobility Panel (GER MP) [20,21]. Table 1 summarizes the parameters that characterize a vehicle availability.


**Table 1.** Vehicle availability characterizing parameters.

The following subsections provide a brief summary of the most important characteristics and differences of the data sets used, as well as a short analysis of the computed vehicle availabilities.

## 3.1.1. Californian Household Travel Survey

Between February 2012 and March 2013, 677 Californian vehicles were equipped with in-vehicle GPS tracking devices. Every trip made by each vehicle was tracked for one week. The publicly available data set contains information about start and end times, start and end location, average speed and miles driven for every trip [22]. In the data set, 19,075 trips by 662 unique vehicles are recorded. The distance traveled varies from less than 0.5 km up to 1289 km. The average distance traveled is 35 km for all conducted trips (see Table 2). Start and end locations are categorized in four categories: HOME, WORK, SCHOOL, and OTHER.


**Table 2.** Mean, maximum, minimum and 95%-ile of distance traveled since last departure from home.

Based on these parameters, the availability of vehicles at 'HOME' can be extracted, with arrival and departure time. Furthermore, the distance traveled is used to calculate the energy used by an average vehicle from its last departure from home. This procedure results in 4062 vehicle availabilities of 592 unique vehicles that contain a vehicle identifier, the distance traveled since the vehicle's last departure from home, its arrival time at and departure time from home. Inconsistencies in the GPS data set, e.g., a vehicle arrives at home but departs for the next trip from another location, are neglected in this analysis.

## 3.1.2. German Mobility Panel

Between September and November 2017, 3867 persons from 1881 households logged their daily mobility behavior in a travel diary. After plausibility checks conducted by the Karlsruhe Institute of Technology (KIT), a total of 70,252 trips were gathered from 1850 persons [21]. The final trip data set includes information about the date, the trip's start and end time, purpose, mode of transport used, duration, distance, and household. In 33,250 of the 70,252 trips logged, the person recorded having driven a vehicle as a driver either as a first, second or third "mode of transport used". In 13,550 of the 33,250 trips, the purpose of the trip was to return home.

Based on the 33,250 trips, a total of 11,458 vehicle availabilities at home are computed by considering household and person identifier, trip purpose, and the trips' chronology.

## 3.1.3. Vehicle Availabilities

In this subsection, the calculated vehicle availabilities are visualized and analyzed. In order to analyze the number of available vehicles at home during an average week, all vehicle availabilities are summed up for each time step of the week. Thereafter, the sums are averaged over all weeks of the field trials. Figure 3 shows the results for the Californian and German data sets. The average number of vehicles available in the Californian data set is 4.5 vehicles and for the German data set 21.3 vehicles. This can be explained by the compressed German field trial period of three months and the higher number of field trial participants (see [20,21]). Despite the differences in quantity, the vehicle availabilities indicate the same trends. At night, the number of vehicles increases until 12 a.m. and then decreases until 12 p.m. This behavior is repeated every day of the week. On weekends, however, the magnitude of the oscillation decreases approximately by a factor of three.

Figure 4 visualizes a histogram of the total number of available vehicles over the distance traveled since their last departure from home. The results of both data sets show an exponential decay with only a few outliers. Ninety-five percent of the American and German vehicles arrive at home with less than 130 km driven since their last departure from home (see Table 2).

**Figure 3.** Percentage of vehicles available at home during an average week. The maximum number of vehicles available at home is 6 for the US CHTS data and 120 for the GER MP.

**Figure 4.** Distribution of the distance traveled since the vehicle's last departure from home—sampled from all modeled vehicle availabilities.

Figure 5 visualizes a relative frequency histogram of the total number of vehicles over the period the vehicles are available at home. Both data sets (US CHTS and GER MP) indicate a periodic behavior with a decreasing amplitude for an increasing time of availability. Most vehicles are available either for less than 1 to 3 h or for 7 to 25 h. Far fewer vehicles are available for 4 to 6 h or for more than 25 h. Table 3 summarizes the mean, maximum, minimum and 95%-ile of the available time for both data sets.

Figure 6 visualizes the distribution of the number of vehicles arriving at home over the hour of the day and the day of the week. Most vehicles arrive at home during the afternoon hours from 3 to 6 p.m. and depart from home between 6 and 9 a.m. Vehicles arrive at and depart from home more frequently during the week than on the weekend. Neither result is surprising considering conventional 9 to 5 working hours.

**Figure 5.** Distribution of the time period vehicles spend at home—sampled from all modeled vehicle availabilities.

**Table 3.** Mean, maximum, minimum, and 95%-ile of time of availability at home.

**Figure 6.** Percentage of vehicle availabilities over (**a**) hour, (**b**) day of arrival time, (**c**) hour, and (**d**) day of departure time.
