By convention the first value is set to NA missing value, because the
difference applies to each sequential pair of locations.
To use this on multiple track ids, use a grouped data frame with tidyverse
code like data %>% group_by(id) %>% mutate(duration = track_time(date))