Automobile manufacturers are making a big push to make fully autonomous vehicles a reality. Whilst fully autonomous vehicles have not really reached most of the general public yet, advancements are always being made, especially around their safety (these advancements also relate to partially autonomous vehicles which are now more common on public roads).
To navigate the roads autonomously and spot signs of danger, these vehicles employ a wide range of sensors. In this article, we look at the different sensors used within autonomous vehicles.
Sensors are an integral part of autonomous vehicles. Without them, it would not be possible for a vehicle to navigate their environment, spot dangers and safely drive along a road. While many sensors are used to measure many different aspects of driving; they generally fall into three categories – cameras and Radio Detection And Ranging (RADAR) and Light Detection And Ranging (LiDAR) sensors.
Sensors not only help to determine the actual environment and the present dangers, they also help the vehicle to provide an appropriate response. These responses can range from accelerating/decelerating to turning, emergency stopping and evasive maneuverers. Whilst the responses are determined by a central software component, it is the sensors that provide the data for such actions to be taken.
Cameras were the first type of sensor to be used in autonomous vehicles and is also the cheapest sensor technology. The cameras used in autonomous vehicles contain a series of CMOS compatible imaging sensors that produce images between 1 and 2 megapixels.
These imaging sensors can be used for a wide range of functions, including providing a precise evaluation of the speed and distance and for determining the presence of objects via their outlines.
Both rear and 360° cameras are employed to provide a comprehensive understand of the nearby environment. Most autonomous cars use 2D cameras, but higher end vehicles have recently started to use 3D cameras.
However, for 3D cameras to provide a realistic representation, 4 to 6 cameras are required which raises the cost of these camera systems and is why they are only used in certain types of vehicles.
For both 2D and 3D cameras to be effective, they need to possess a dynamic range of more than 130 dB, with some of the higher end cameras possessing a dynamic range of 145 dB. Additionally, cameras used in autonomous vehicles need to have a single to noise ratio close to 1 and a high frame rate.
Both short range (24-GHz) and long range (77 GHz) RADAR are employed at the front and the rear of autonomous vehicles and their primary use is to monitor the surrounding traffic. Because these vehicles use both short and long-range sensors, they can now monitor traffic that is within a few centimeters of the vehicle and vehicles which are hundreds of meters away.
Both short and long-range RADAR detect objects in the surrounding environment by emitting electromagnetic radio waves and detecting their return by a receiver. Both RADAR range types have advantages, hence they are both employed within the same vehicle. Long-range RADAR systems are useful because they provide more accurate and precise measurements for speed, distance and angular resolution. On the other hand, short-range sensors are much smaller in size and suffer less from interference problems.
The combination of these systems is crucial for any autonomous vehicle, as they can help with blind spot monitoring, changing lanes, rear-end collision warnings, parking, cross-traffic monitoring, braking (including emergency braking) and automatic distance control.
LiDAR sensors are the newest addition to autonomous vehicles and work in a similar way to RADAR. Where they differ to RADAR is that they use a transmitted and received laser signal rather than radio waves.
LiDAR sensors that are used for general automobiles (and not military vehicles) rely on an infrared (IR) laser. LiDAR IR systems can either use a rotating laser or a solid-state laser, both of which are aided by the use of microelectromechanical system (MEMS) components.
Single-photon avalanche diodes (SPAD) are often used to detect the reflected photons. Once a photon reaches the diode, the semiconducting p-n junction within the MEMS components causes the current in the diode to breakdown, which is then detected and outputted. To achieve the required sensitivity to make these systems reliable in autonomous vehicles, an array of diodes is often used to detect the inbound photons.
LiDAR sensors are used within autonomous vehicles to measure the distance between stationary and moving objects and to create a 3D image of the detected objects; and can measure said objects from up to 100 meters away.