Connected Cars

How to develop autonomous driving solutions

It’s hard not to be excited about the potential of the autonomous car. But as developers know, building a car that can learn to navigate through its environment takes more than raw processing power. Intel can help you develop for autonomous cars. With deep learning you can create an accident-free experience with cars that can sense, learn and make proper decisions.

Achieving this means greater reliance on sensors, data and processing power. And that’s true whether you’re developing for highly automated driving in which the driver is supported in functions such as navigation, security and image recognition or whether you’re working on fully automated driving (FAD) where the driver is essentially just a passenger.

The building blocks of an automated vehicle

A deep learning foundation for advanced driver assistance systems (ADAS) is an essential component of a safe autonomous vehicle, but there are many other attributes to consider, including:

  • The ability to process enormous amounts of environmental data
  • The ability to sense surroundings, including:
    • Creating access to hardware accelerators for developing common computer vision routines
    • Providing inference data gathered from cameras
    • Methods to extract features from video and track them

Technology inside the car… and out

It’s important to remember that a range of technologies both inside and outside the car are responsible for powering autonomous vehicles. In particular:

  • Data Centre: One autonomous car generates a volume of data equivalent to almost 3,000 people, making the data centre critical for storing, sharing and protecting the data generated from deep learning algorithms
  • 5G: 5G connectivity delivers ultra-low latency at gigabit speeds and high bandwidth. This allows for intelligent and agile networks to give priority to the safety-first devices needed for an automated vehicle
  • Human Machine Interface: That means a software-defined cockpit that consolidates cluster displays with in-vehicle infotainment systems. The cockpit should include media management, security systems and cloud connections. Ultimately, this design should build trust between the driver and vehicle

With automated vehicles set to rely more on sensors, data, and processing power, it’s important not to overlook the importance of in-vehicle computing. Intel’s portfolio of power-efficient processors, field-programmable gate arrays (FPGAs), and software are designed to deliver high-compute performance per watt.

Get started today

Request access to the Intel GO Automotive SDK Beta to enjoy a comprehensive toolkit to build high-performing, power-efficient designs for in-vehicle and cloud-based data centre platforms.

If you are a data scientist, system designer or developer of autonomous driving solutions, this SDK will help you to maximise hardware performance, optimise systems and applications and advance perception sensor and deep learning algorithms. The SDK includes several workflow modules and optimisation tools, some of which are specialised for automotive development.

Getting the right hardware and software in place now will enable developers to build safer, smarter autonomous vehicles. Visit the Intel® Developer Zone to view the full range of tips and tools available to support emerging autonomous driving capabilities and help developers create their own driving experiences.

Click to comment

You must be logged in to post a comment Login

Leave a Reply

To Top