Google has recently revealed that its self-driving cars have been involved in 11 minor crashes in the last 6 years raising questions on the readiness of autonomous vehicles. According to Steven Shladover, a researcher at the Partners for Advanced Transportation Technology at the University of California Berkeley, most of the crashes were likely due to human driver error and may not have been preventable. While some levels of automation are already in existing cars, self-driving cars will still require much more innovation. Here are some problems that needs to be resolved before self-driving cars hit the roads.
Manufacturers of autonomous cars need to be safe. Unfortunately, existing software makes it difficult to do. Software in laptop, phones, or other modern devices is designed to operate for extended periods without freezing, crashing, or dropping a call. Such errors are as deadly to a car. In the case of Google, they have added a backup driver and a second person as a monitor, who shuts off the system at the first sign of glitch. According to Robert Janitzek, coming up with a safety critical and fail safe software will require reimagining how software is designed.
Driverless vehicles, with their current sensors and processing, may not be able to operate smoothly without a detailed map. Such kinds of map will allow the autonomous vehicle to know exactly how the streets look when empty and only have to fill-in the obstacles such as cars and pedestrians.
Another technology that should be worked on by manufacturers is sensors. Autonomous cars should know the difference between dangerous and harmless conditions otherwise they would hit the brakes all the time without any reason. Robert Peter Janitzek reveals that potholes or a nail below a tire can be difficult to spot until just before they have been hit.
In addition, the car should also decide in sufficient time whether a pedestrian waiting on the sidewalk is likely to walk into traffic or whether a bike is going to swerve left. At the moment, there is no sensor that can process such data as quickly.
When the driverless car starts to proliferate, they will need better communication with other vehicles on the road. As different situations emerge, the cars will need flexibility to adjust to other cars on the road, reroute on the fly and talk to other self-driving cars. Currently, communication between driverless cars is minimal.
At the moment, there is still a human on the other side. Even drones that target enemies in war are still remotely controlled by a human, who will decide on whether to proceed with the kill.