Given the correct inputs, there are many tasks that robots (or machines or driverless cars) can do much better than people. The keyword here might be 'input' however.
In the case of driverless cars, they are using many sensors, cameras and radar to control those input but they are not perfect just yet. In time however, we can expect few traffic accidents, due to an autonomous system's increased reliability and faster reaction time compared to human drivers.
Remember, Google self-driven cars have managed over 700,000 autonomous miles without any collisions. We can expect increased roadway capacity and reduced traffic congestion. The reduction in congestion comes about because of the reduced need for safety gaps and being able to better manage traffic flow.
Some autonomous truck vehicles are being used on a daily basis in some sectors such as mining. For instance, since December 2008, Rio Tinto Alcan has been testing the Komatsu Autonomous Haulage System which is the world's first commercial autonomous mining haulage system in the Pilbara iron ore mine in Western Australia. Rio Tinto has reported benefits in health, safety, and productivity. In November 2011, Rio Tinto signed a deal to greatly expand its fleet of driverless trucks. Other autonomous mining systems include Sandvik Automine's underground loaders and Caterpillar Inc.'s autonomous hauling.
Daimler are developing the Mercedes-Benz Future Truck 2025 in the belief that drivers of the vehicle will be "transport managers" rather than truck drivers. The vision for these driverless trucks is that they start, and organise their collection, transport and delivery at their destination autonomously. It is hoped that they will react flexibly to spontaneous situations like traffic jams by negotiating new delivery times with customers. Nissan and other vehicle companies are also testing similar systems.
As crash avoidance technology becomes more common, you can be sure that things like lane departure warning will become mandatory. Previous attempts to commercialise automatic vehicle control have failed because they required dedicated infrastructure and vehicles that must remain entirely under automatic control.
New technologies, like CMOS radar-on-a-chip and all-weather LIDAR, will lead to more intelligent and safer vehicles.
Dr Kevin Curran, Senior member of the IEEE