Note to Google: Self-Driving Cars Need More Than Self-Control
Interested in operating a self-driving car? You’re going to need some training for that, the National Highway Traffic Safety Administration said Thursday. In its first response to the emerging automated vehicle field, the agency released a list of guidelines, titled the “Preliminary Statement of Policy Concerning Automated Vehicles” that details how states should go about handling the emerging technology of self-driving vehicles. One of the top recommendations includes requiring operators of self-driving cars to gain extra training and special licenses, proving their competency on public roads.
As both established automakers and new faces, including Google Inc. (NASDAQ:GOOG), are beginning to make rapid developments in the automated vehicle field, the NHTSA is eager to to propose guidelines and policies that require potential drivers to research and train in self-driving cars before operating the vehicles on public roads. Though the self-driving cars may be effective in reducing highway accidents and fossil-fuel emissions, they can be dangerous for drivers inexperienced with automatic vehicles.
Thus, the NHTSA recommends that states mandate potential self-driving car operators to undergo training, pass a test, and obtain certification that proves the driver’s operation of the vehicle for a certain minimum of hours before hitting the public roads.
The increased production of self-driving vehicle technology both spurs concern and alleviates it. The NHTSA has addressed the dangers of removing the driver from direct control of a car, and explained in a statement Thursday that “it is considering rules that could effectively require car makers to equip vehicles with technology that automatically applies the brakes when the car senses an imminent crash.”
However, the upside to self-driving car technology? Diminishing the number of bad drivers on the road. Even though the NHTSA is concerned that removing the driver from control of a car would inhibit his ability to press the breaks before a crash, a recent study found that only 1 percent of drivers in the sample applied the brakes full force before a collision.
Another upside of the emerging technology lies not only in its potential to reduce carbon-dioxide emissions, but also in the cost-saving opportunities and potential profit possibilities it affords companies, such as Google, The Wall Street Journal explains, which has drivers operating cars for the purpose of collecting images and data for its Street View mapping product. The company has also suggested it could “sell automakers the mapping and machine-vision technology it’s developing for its self-driving cars, or offer that technology for free as part of a package of other Google products and services.”
Unfortunately, Google is still in the midst of working on its self-driving car’s glitches, as elucidated in a report by NBC in March that detailed the vehicle’s difficulties. The self-driving cars have trouble navigating in the snow, because they have a hard time seeing objects when snow is on the road obstructing their view. Additional problems include the dependence of the cars on computer maps that breed issues when the car is lost, or is trying to navigate a construction site or accident. Google maintains that it is working through its flaws so as to release the vehicles to the public market as quickly as possible.
And forecasts for autonomous driving show that this is just the beginning. Current suppliers are working to further expand the capabilities of automated vehicles, including features that warn against collision and offer more advanced cruise control and electronic steering options. Germany’s Continental AG is even working with car companies to develop certain lane-keeping systems that allow vehicles to drive themselves under certain conditions, such as congested roads.
Industry executives expect vehicles with self-driving capability to be available by 2020, and fully autonomous vehicles by 2025. Needless to say, the NHTSA is about to have its work cut out.