Can I use self-driving car's data set for left-hand drive cars which drive on the right lane for right-hand cars which drive on the left lane?

Question

Can I use self-driving car's data set for left-hand drive cars which drive on the right lane for right-hand self-driving cars which drive on the left lane?

Answer 1

No

Various reasons:

• From Wikipedia: "Left- and right-hand traffic":

  "In most countries, rail traffic follows the handedness of the roads, although many of the countries that switched road traffic from LHT to RHT did not switch their trains. Boat traffic on rivers is effectively RHT. Boats are traditionally piloted from the starboard side to facilitate priority to the right.".

  Sometimes cars and trains interact, especially at rail crossings, a frequent source of accidents.

• From Wikipedia: "Priority to the right":

  "Some countries use the priority-to-the-right rule, despite driving on the left. Australia uses the priority-to-the-right rule at four-way intersections where the roads all have equal priority, but specific rules apply for T-intersections. Singapore also uses priority-to-the-right, as well as priority to vehicles going straight and turning vehicles to give way to vehicles going straight.".

• Wikipedia: (local legal exceptions) "Boulevard rule":

  "Maryland is among the U.S. states which follow this rule, but not all states have similar provisions in statutes or case law. New York applies the rule to traffic entering public roads from private driveways or alleys, but not where public roads intersect.".

• Wikipedia: Changing sides at borders":

  "Traffic Switchover sign at the Thai–Lao Friendship Bridge. Although many LHT jurisdictions are on islands, there are cases where vehicles may be driven from LHT across a border into a RHT area.". See the Lotus Bridge in Google Maps.

• Vehicle regulations: DriveSmartBC Article: "Right Hand Drive Vehicles in a Left Hand Drive World":

  Some areas allow RHD vehicles, despite a majority of the vehicles being LHD and the roads being designed for same.

That is an incomplete list of exceptions.

See also: "Advantages and disadvantages between left hand and right hand units" for a brief list of pros and cons, which points out some additional exceptions.

Answer 2

I would suggest to look into dataset like nuscenes, which have both left and right driving data. So if you want to test, then you can use left driving dataset for train and later right driving for testing or vice versa. I believe that will be a better option.