How Self Driving Cars Work with Embedded Systems

a Self-driving car going down the road in a heavily forrested area

We are all aware that self-driving cars are the transportation of the future. Google’s self-driving cars have logged almost 2 million miles. Other automakers, such as Volvo, Kia, BMW and General Motors are also working to develop fully automated cars. These cars will only be possible through the work of multiple embedded systems.

Embedded systems are computer systems (CPUs that contain a suite of sensors) that are developed and programmed for a specific task. Unlike other computers, where the coding can be changed or the operating system can be customized, an embedded system is created to receive information and then do the task it was designed to do based on that information. Embedded systems are used in various industries, but perhaps there is no application that is as interesting as in how it is used in transportation.

Embedded Systems and Self-Driving Cars

Self-driving cars need to have all the necessary technology to make these vehicles fully aware of their environment and able to react to changes in their environment. These cars need to understand the safest way to react to pedestrians, other cars on the road, road obstructions, and other driving risks. They also need to consider changes in weather. For example, if there is a sudden rainstorm or snowstorm, a self-driving car needs to know exactly how to drive safely in that environment.

The only way  a self-driving car will be able to receive and properly react to  information is through embedded systems. Your car already contains hundreds of embedded systems that you may not even be aware of. The climate control, the smog monitor, built-in safety systems - are all controlled by embedded systems. There are very advanced embedded systems in some cars that are very similar to what will be used in self-driving cars such as Ford’s adaptive cruise control (ACC) system that helps cars keep safe distances from the cars in front of them. However when it comes to self-driving cars, there are specific embedded systems that will be absolutely crucial for their success.

Important Embedded Systems for Self-Driving Cars

While self-driving cars will need all of the same embedded systems that your car currently has, there are specific systems that will be much more important, absolutely vital for self-driving vehicles.

1.Global Positioning Systems (GPS)

Since there will be no human driver, the self-driving car will need to think for itself as it travels between locations. For a self-driving car to get from its starting point to its destination, it will need to have GPS. GPS relies on signals that it receives from at least four GPS satellites. These signals allow the car to know its location, as well as its speed and direction. This information will help the car to drive at a safe speed, relative to the speed of the cars around it.

Many GPS devices are becoming much more accurate, and can also monitor traffic and discovering faster routes to save time for the passengers. However, the difficulty is that sometimes GPS are vulnerable to radio interference, or can be blocked or distorted by skyscrapers and other factors in the environment. There are other embedded systems being developed for navigation, such as automotive inertial measurement systems - gyroscopes, accelerometers and other instruments that supplement the GPS. Together, they ensure  the car will always know where it currently is and know exactly how to get to where the passengers need to go.

2. Radar Systems

Radar is already used in many cars that have ACC embedded systems or blind spot monitoring systems. The radar in self-driving cars will be much more comprehensive. Self-driving cars will need to have front-facing and rear-facing radar systems to provide the car with the information about the objects around the vehicle. It will inform the car when an object is too close so that the car will be able to react accordingly. The 24GHz radar, often placed in or near the front and rear bumpers, is already being used in cars for short and mid-range collision avoidance, self-parking, and blind spot detection. However, in self-driving cars, the radar systems will need to be more advanced. Automakers will need to adopt 77GHz radars that will be able to detect obstructions in a longer range.

3. Forward-Looking Cameras and Other Sensors

Forward-looking cameras as well as radar will work together to give the car more information about its surroundings. Not only will it help inform the car of any obstructions, but it will also inform it about any traffic signs or lights, junctions, and any other information that it will need to know in order to correctly interact with any other vehicles on the road. Radar itself would not be able to detect this type of information, which is why automakers will need to rely on developing forward-looking cameras that will gather this information and other embedded systems will calculate exactly what the car should do based on this information.

Google has also uses other sensors such as light detection and ranging (LIDAR), which can construct 3D images and calculate the range to objects that are further down the road. The current Tesla autopilot uses sonar to detect any objects that are 16 feet away - close to the vehicle. These sensors, with the radar, will all be working together to create an accurate depiction of the environment around the vehicle. The more information that is given to the vehicles, the safer the ride will be for the passengers.

4. Digitally Controlled, Highly Precise Braking Systems

When it comes to safety, a braking system is extremely important. Unlike the other embedded systems that  gather information, this embedded safety system reacts to the information that it has been given. Your braking system will need to take into account any obstructions on the road. It will also need to take into account any hazardous road conditions that are caused by weather changes. While cars that use ACC already have digitally controlled braking systems, the systems in self-driving cars will need to be highly precise and be able to break in accordance with all the information that the rest of the embedded systems have given it.

The Future of Embedded Systems in Self-Driving Cars

There are so many more embedded systems that will need to be used in self-driving cars there will need to be systems that handle steering and speed regulation as well. Embedded systems often take a long time to develop since early prototypes will have imperfections. As technology advances, they will continue to create embedded systems that will improve the ride for the passengers. Engineers are already working to making these embedded systems to work together to make self-driving cars are fuel efficient as possible.

As more and more automakers begin developing more self-driving cars, engineers will have the ability to create embedded systems that will affect how the self-driving cars interact with each other. This will make driving safer and more efficient for everyone. These interactions could possibly include the ability to have cars that are able to drive a particular distance away from each other in a way that reduces drag and is more aerodynamic. This will that more energy efficient for all the cars. Also, they will be able to communicate with other cars, so that one car can inform the other that it will need to brake ahead. Scientists are researching strategies that cars can use to work together to help everyone get to the destination faster and engineers will be able to develop technology with these strategies in mind. Engineers constantly need to adapt and continue to experiment in order to perfect their embedded systems for self-driving cars.

Total Phase offers products to help any embedded systems engineer create functional systems for the automotive industry, such as the Komodo CAN Duo Interface. Click below if you would like to learn more about embedded systems in automobiles and how they can be monitored as well as tested, or have a personal demo specifically to address your needs.

Request a Demo