How are autonomous vehicle sensors changing the ADAS systems?

Autonomous vehicles are one of the most complex techniques that define the way we drive our vehicles. Autonomous vehicle sensor help to bring a powerful combination of electronics and innovative software to deliver quality results. 

How do autonomous vehicle sensors bring the best results?

Intricate autonomous vehicle sensors and complex software algorithms help to control autonomous vehicles and their functioning. The quality of these sensors determines the level of automation needed. 

ADAS systems with LIDAR technology bring advanced results in self-driving technology. This is used to determine and generate high-resolution images of the target objects. The use of lasers helps to evaluate the right distance between objects and navigate through the complex settings of the real world. 

The LIDAR systems adopt a noncontact range-finding method using which short pulsed lights are emitted. The target and backscattered pulse are spotted and then processed by the software to determine the distance between the target and the camera.

LIDAR systems also help to obtain a real-time representation of the surroundings where the 3D geometry (volume, distance, size) is recognized. Hence, the technology helps to ensure the safe and secure navigation of autonomous vehicles and prevents clashes. LIDAR systems help to create the right images and high-resolution maps along with vegetation. 

How do the LIDAR sensors work?

Imaging LIDARs use a time of flight principle. This assesses the distance between the target and light sources. By calculating the delays between the light emission and return, it is easy to form a 3D point cloud. 

This can be accomplished by modifying the signal power and/or frequency. By calculating the time required for the signal to arrive back to the camera in a number of points one gets the pont cloud of the scene. To determine the distance between the target and LIDAR a pulsed approach is used in most cases. An elementary measurement approach involves emitting short and powerful light pulses toward the target, and measuring the time taken by the echo to reach out to the detectors with an electronic time counter. 

In this technique, a small portion of the outgoing pulse is directed toward the receptor to start the timer. The rest of the pulse is transmitted in the direction of the target. The timer stops counting when the pulse reaches back to the receptor. 

The time calculated is multiplied using the speed of the light in a given medium. Since the speed of the light is constant in the medium and the distance is directly related to the time traveled. This is then doubled to match the specifications as per target distance (light goes forth and back to the target).

When a light pulse is emitted and directed to the target towards the LIDAR. After reflection, the amount of light that is recovered by the same is dependent on factors like – 

• Transmitted light intensity,
• How far the target is,
• Size of the target,
• Size of the optics, etc..

For the sensors to work effectively, the pulse received back by the detectors should be more than the minimum noise intensity. Hence the power of the transmitted pulse decides the maximum distance the LIDAR can measure, which has to be kept within eye safety limits defined by normative.

Components of the LIDAR sensors 

A typical LIDAR system can have the following components – 

• Emitters for emitting light waves into the environment, typically lasers. These help to scan and measure the time needed to reflect from objects to judge the actual distance and terrain depth. 
• The receiver helps to capture the reflected light waves. This helps to understand the form, size, and volume of an object.
• Signal processing system to ensure effective results.

Such techniques are also accurate with depth perception. It also helps to control the vehicle in terms of obstacles and control its performance during night or challenging weather.

To increase the distance, the pulse intensity needs to be increased proportionally. This can however be increased to a particular level. The maximum light intensity is capped by the average power that is safe for the human eye. The pulsed approach is quite simple and has a low signal-to-noise ratio. 

The data collected by the system is then scanned through the scene and used to create a 3D image. The same helps to measure the LIDAR and create virtual images of the entire terrain. With autonomous driving systems, the collected data is used to track and monitor performance in real time. 

Difference between LIDAR, RADAR, and camera systems 

LIDAR systems offer a complete perception of the surroundings. Unlike the limited reliability or the conventional 2D systems, these products can create a strong impact. However, the performance of the LIDAR depends on the environmental factors. In challenging weather the RADAR may perform better, but always with very limited number of points in the scene. 

Thus, RADAR cannot create a detailed image of the surroundings as long as it only collects a very limited number of points compared to LIDAR. When combined with the LIDAR, it can help boost efficiency. This also helps to dramatically control the AV in terms of speed. Data fusion of LIDAR and RADAR is a critical endeavor to compensate for failure modes of both systems.

Further, through LIDAR and radars, it is easy to meet the critical functions of the ADAS. Still, the systems must be connected to the camera. Despite they do not see geometry, but only its projection on a plane, they allow to detect traffic lights, traffic signs, etc. Cameras offer HD detail but can struggle with elements like glare.