Building the Foundation for Self-Driving Cars

Navigation and Sensors

Just as Rome wasn’t built in a day, self-driving cars are not composed of one singular innovation or technology that evolved overnight. They are a culmination of innovative ideas and products that are constantly evolving to contribute to even greater advancement.

When we look at the levels of automation already achieved in todays’ vehicles, we see the foundation for completely self-driving vehicles. Two important building blocks in that foundation both involve driver-assistive: navigation and mapping tools and sensors.

Mapping and Navigating the Road

The evolution of our in-vehicle navigation, from printed maps to targeted tracking, has changed rapidly in the last decade. Despite the overwhelming usefulness of personal navigation devices in terms of finding your way, the original intention of the Global Positioning System (GPS) was solely for military purposes.

This initial system, NAVSTAR, was developed by the United States Department of Defense in 1973, and officially took off five years later when the first four satellites were launched into space. It took almost another 20 years for GPS to reach full operational capacity with 24 working satellites, and then another 15 years to sign legislation that made that full coverage available to consumers.

Although the initial concept of GPS became valuable to driver navigation, the technology was not localized enough to direct a self-driving car. Expanding on that concept, companies such as Ford Motor Company have developed more advanced implementations of the technology to increase its accuracy. 

LiDAR, used by the Ford Fusion Hybrid, is much more accurate than GPS, pinpointing the location of the vehicle within a centimeter. LiDar works by emmitting short pulses of laser light that allows the vehicle to create a high-definition 3D image of what’s around it. The technology enhances a mapping tool with sensors to better identify the car’s surroundings.

Localization allows the car to “know” its position relative to the world and environment around it. The more precise the localization, the safer and more reliable a self-driving car becomes.

These navigational and mapping technologies can also improve the car’s performance in inclement weather. In the case of the Ford Fusion, when LiDAR and other sensors such as cameras can’t see the road, high-resolutions 3D maps created from a test environment detect above-ground landmarks to pinpoint the car’s placement on the map, thus allowing the vehicle to localize despite inclement weather.

Sensing the Cars' Surroundings

Sensors are responsible for evaluating an event or action, collecting data, and determining the appropriate action or response. In this way, they play off the other in-vehicle technologies, such as GPS, to collect and interpret the data that is being presented.

Much like GPS, sensors have been in cars for years, with different levels of complexity. According to research from CTA and Vision Systems Intelligence, sensors are “necessary for passive safety systems like rear camera or proximity sensors and in many countries they are required by regulation,” but they also provide extensive opportunities for increasing the car’s capability. “Sensors have to create a digital representation of the environment, plus have the logic to understand the scenes and convert this to control signals.”

To achieve the complexity of full 360-degree coverage, many types of sensors are required for long-range monitoring and quick reacting.

Moving Forward

Through the implementation of GPS and mapping technologies and sensors, we continually lower the risk of driver error, which causes 94 percent of car crashes in the United States. By employing technologies that compensate for driver distraction or error, we greatly lower the risk of dangerous crashes.

Although the accuracy of these technologies is constantly improving, there is still a long road ahead before they can be used cohesively to enable a fully self-driving car. Obstacles such as inclement weather and high production costs make it harder to achieve autonomy, but examples such as LiDAR that combine the powers of sensors and mapping are moving us closer to that reality.