Using Deep Learning, Driverless Cars Will Soon Start Learning How To Interact With The Environment

The phenomenon of driverless cars is taking the automobile industry by storm. In fact, scientists and software engineers are still trying to decide how they should make the driverless cars more intelligent. Perhaps, a California startup, has the answer. The new company will use deep learning as part of ongoing tests to make driverless car learn from their environment.

Deep learning is an advanced form of artificial intelligence. Actually, its not new because researchers at the Stanford University are already using it to test different technology. However, this is the first time that a company will use deep learning as a tool to train driverless cars.

Interestingly, the company only reveled its plan to test the driverless vehicles because it had to get the test license from the Government of California. According to sources, the license will allow to use driverless car for the training purposes. Accordingly, the company has recently hired high-profile automobile executives that includes former General Motors director Steve Girsky. It is also able to obtain nearly $12 million in funding, which indicates future potential.

According to president and co-founder Carol Reiley, deep learning in automobiles will work by allowing the care to decide from millions of test cases. For instance, engineers can offer thousands of different scenarios for the driverless car to learn in matter of days. The amount of and data that modern computers are able to store will help the computer to evolve with the passage of time. As more unique circumstances are presented, the artificial intelligence will reach new levels of coordination. Carol also revealed that deep learning is not limited to driving the car. Instead, the technology will enable the computer to make instant decision on what is best for other drivers, pedestrians and environment. Overall, it means that is building a computer that will not only save the driver, but is also concerned about the safety of others. If true, other companies can also learn from

As of now, the deep learning will continue to be in a Beta mode. As co-founder of the company, Sameep Tandon, sees it, deep learning requires continuous learning process. The learning cycle continues because it is impossible to make a guide for a computer that is always facing new scenarios.