The company said that the new update brings improved object detection network architecture for non-VRUs (e.g. cars, trucks, buses), seven per cent higher recall, 16 per cent lower depth error and 21 per cent lower velocity error for crossing vehicles.

It also brings a new visibility network with 18.5 per cent less mean relative error, the report said.

The company also said that it allows more room for longitudinal alignment during merges by incorporating modelling of merge region end.

The software enables the vehicle to drive autonomously to a destination entered in the car's navigation system, but the driver needs to remain vigilant and ready to take control at all times.